Professor Greville Corbett

As part of the attempt to understand the linguistic origin and cognitive nature of grammatical gender, we designed six psycholinguistic experiments for our language sample from Vanuatu (Merei, Lewo, Vatlongos, North Ambrym) and New Caledonia (Nêlêmwa, Iaai). Each language differs in number of classifiers, and whether nouns can freely occur with different classifiers, or are restricted to just one classifier (similar to grammatical gender).. Free-listing: participants heard a possessive classifier and listed associated nouns. This revealed the different semantic domains of classifiers, the salient nouns associated with each classifier, and showed whether participants listed the same noun with different classifiers.. Card-sorting: Participants free-sorted sixty images, followed by a structured sort according to which classifier they used with each picture. We compared whether similar piles were made across sorting tasks to reveal whether the linguistic classification system provides a structure for general cognition.. Video-vignettes: Participants described 24 video clips which showed different interactions between an actor and their possession, evoking a classifier. This tested both typical and atypical interactions to see if the same or different classifiers were used.. Possessive-labelling: Participants heard 140 nouns in their language and responded by saying the item belonged to them, which meant using a classifier. This measured* inter-speaker variation in the use of classifiers for particular items, reaction times and inter-speaker variation for different possessions.. Storyboards: eight four-picture storyboards were presented to participants. We recorded participant responses, uncovering if the same classifier was used in consecutive parts of the larger story and whether the classifiers were used anaphorically.. Eye-tracking: eight line-drawn pictures were combined in a paired-preference design. An eye tracker recorded fixation times. Participants heard the auditory cue of a classifier before being presented with a pair of images. This provided objective measures of automatic processing to identify patterns in attention.

We undertook a vernacular literacy education survey with teachers from three language communities in Vanuatu. Our main objective was to understand more about the experiences of using the local vernacular language as the Medium of Instruction (MOI). Though this research is a pilot study, the results are intended to inform current education policy. In particular the study’s main aims are to understand:1. Teachers’ perspectives on speaking and using the main community language in the classroom and what training they require.2. How many children speak the main community language in each community and their enjoyment of using the language in the classroom.3. What further teaching resources are essential for effectively using the community language as the MOI.We wish to thank Helen Tamtam from the Curriculum Development Unit and Prof. Robert Early from the University of the South Pacific for their collaboration in developing the survey. We wish to acknowledge the help given from the three language communities of North Ambrym, Lewo and Merei, in particular, for the support from Yanick Tekon, Simeon Ben and Adam Pike. Furthermore, we wish to thank Mabel Baldry and Emily Marsh for the infographics. Finally, thank you to Penny Everson for her help in preparing the document.

This document is the literacy, language and teacher training survey that was conducted with schools from the Merei, North Ambrym and Vatlongos language communities in Vanuatu in June/July 2023. The survey was conducted as part of the project on Sustainable Endangered Language Maintenance in the South Pacific, funded by the University of Surrey’s Economic and Social Research Council (ESRC) Impact Acceleration Account.

Possession in Lewo: grammar and dictionary.A pedagogical grammar of possessive constructions in Lewo (a language of Vanuatu). The grammar is written in Bislama, English and French. The dictionary shows which nouns occur with which possessive constructiuons and which nouns occur with each possessive classifier, informed by a free-listing experiment.Ilustrated by Isabelle Ritzenthaler.

Possession in Rral: grammar and dictionary.A pedagogical grammar of possessive constructions in Rral (also known as North Ambrym, a language of Vanuatu). The grammar is written in Bislama, English and French. The dictionary shows which nouns occur with which possessive constructiuons and which nouns occur with each possessive classifier, informed by a free-listing experiment.Ilustrated by Isabelle Ritzenthaler.

A creative 1 minute video about the optimal categorisation project and our community outputs for a general audience. Created for the Nature science in shorts video competition: https://www.nature.com/immersive/scienceinshorts/index.html.

Possession in Vatlongos: grammar and dictionary.A pedagogical grammar of possessive constructions in Vatlongos (also known as South East Ambrym, a language of Vanuatu). The grammar is written in Bislama, English and French. The dictionary shows which nouns occur with which possessive constructiuons and which nouns occur with each possessive classifier, informed by a free-listing experiment.Ilustrated by Isabelle Ritzenthaler.

Possession in Iaai: grammar and lexicon.A pedagogical grammar of possessive constructions in Iaai (a language of New Caledonia). The grammar is written in English and French. The dictionary shows which nouns occur with which possessive constructiuons and which nouns occur with each possessive classifier, informed by a free-listing experiment.Ilustrated by Isabelle Ritzenthaler.

Happy Families Card Game of Possession in the Iaai LanguageAn innovative card game based on happy families / jeu de sept familles where players collect the 'family' of nouns that go with a specific possessive classifier. With instructions in Fench and Iaai.Cards designed and illustrated by Isabelle Ritzenthaler

Gender is a feature of special interest because it provides a dramatic demonstration of just how different languages can be. For many of the Indo-European languages it is an important part of the grammar that is realized in a high proportion of utterances. In most Dagestanian languages, such as Tsakhur or Archi, it is still more salient: it suffuses the syntax and morphology, appearing on some unlikely agreement targets. In almost all the languages of the Austronesian family, it is simply missing. This article first examines the definitions of gender to ensure that we are comparing like with like. There has been careful research to do this, and it will become clear that within the languages that do have gender there is a considerable variety of systems. This is particularly apparent in the ways in which speakers assign nouns to genders. Gender systems may have sex as a component, as in languages with masculine and feminine genders; but, equally, sex may be irrelevant, as in the Algonquian languages, where the distinction is between animate and inanimate. This article outlines the distribution of gender in the world's languages, and reports on recent developments in the field.

Imenice pluralia tantum su fascinantna pojava, za koju se slavisti interesuju još od Braunove teze (1930), pa i ranije. Za to postoji nekoliko razloga. Prvo, mnoge od ovih imenica su defektne jer, iako su brojive, nemaju jedninu. Drugo, dok defektne imenice obično uključuju sporadične praznine u promeni (kao što je to, po jezičkom osećanju mnogih govornika, slučaj sa genitivom množine ruske imenice *mečt), klase imenica pluralia tantum često su – bar delimično – semantički predvidljive (tj. ove imenice su podložne „uopštavanjima srednje veličine”, Koenig 1999). Međutim, ova predvidljivost je ograničena: up. u ruskom imenicu binokl ‘dvogled’, koja ima standardnu promenu, sa imenicom sani ‘sanke’, koja je imenica plurale tantum. Treće, da bi se mogle upotrebiti kao brojive, imenice pluralia tantum često zahtevaju neko dodatno sredstvo, npr. zbirne brojeve. I u opštoj lingvistici postoji stalno interesovanje za ove imenice (npr. Visnievski 2009). Zato je glavni cilj ovog referata da pokaže poseban značaj podataka iz slovenskih jezika za opštelingvističko ispitivanje imenica pluralia tantum. Taj značaj je dvostruk. Prvo, kod nekih klasa ovih imenica se uočavaju zanimljive razlike u različitim slovenskim jezicima. I drugo, treba preispitati tvrdnju da je, iz semantičkih razloga, prirodno da se predmeti koji se javljaju u paru označavaju imenicama pluralia tantum, s obzirom da se ovakvi predmeti označavaju imenicama pluralia tantum i u jezicima koji imaju dual (kao u slovenačkom i gornjem i donjem lužičkom). Uz to, u ovim jezicima se imenice pluralia tantum mogu koristiti za jedan, dva ili više od dva referenta (kao u donjem lužičkom, Janaš 1976/1984). Dakle, ove imenice zaslužuju našu stalnu pažnju bilo da je reč o razumevanju broja u slovenskim jezicima ili o opštelingvističkim aspektima imenica pluralia tantum. = = = Pluralia tantum nouns are indeed fascinating, and Slavists’ interest in them goes back to Braun’s thesis (1930) and earlier. There are several reasons for this. First, many of these nouns are defective, since they are countable yet they lack a singular. Second, while defectives typically involve sporadic gaps (as with Russian genitive plural *mečt for many speakers), sets of pluralia tantum nouns are often semantically predictable, at least in part (they are subject to ‘middle-size generalizations’, Koenig 1999). This predictability is limited, however: compare Russian binokl ‘binoculars’ (which is a normal noun) with sani ‘sleigh’ a plurale tantum noun. Third they often require some sort of repair (such as the use of collective numerals) so that they can be treated as count nouns. Pluralia tantum nouns are of continued interest in the general linguistic literature (see, for instance, Wisniewski 2009), hence it is timely to consider the particular interest and contribution of the Slavonic data, the focus of this paper. Two aspects stand out. First the Slavonic languages provide semantic classes of these nouns, which vary across the family in interesting ways. And second, the claim for the semantic naturalness of paired objects being pluralia tantum needs reassessing, given that these nouns are pluralia tantum even when the dual is available (as in Slovene and Upper and Lower Sorbian. Moreover, in these languages pluralia tantum nouns can be used for one, two, or more than two referents (as in Lower Sorbian, Janaš 1976/1984). Thus these nouns deserve our continued attention, both for Slavonic internal and for general linguistic reasons.

The Mian and Kilivila Collection contains information pertaining to the nominal classification systems of two indigenous languages of Papua New Guinea, Mian and Kilivila. Kilivila has a single system of classifiers, with a great number of distinctions, while Mian has a dual system, which combines four genders and six classifiers. The Digital Collection on this website permits users to gain a greater understanding of these systems by exploring images of Mian and Kilivila objects and people. Users are also able to test what they have learnt about the classifications systems of these two languages by taking the online Quiz.

The Mian and Kilivila Collection contains information pertaining to the nominal classification systems of two indigenous languages of Papua New Guinea, Mian and Kilivila. Kilivila has a single system of classifiers, with a great number of distinctions, while Mian has a dual system, which combines four genders and six classifiers. The Digital Collection on this website permits users to gain a greater understanding of these systems by exploring images of Mian and Kilivila objects and people. Users are also able to test what they have learnt about the classifications systems of these two languages by taking the online Quiz.

A general typology of number systems has to confront the problem of variation both in the number values in different languages and in the inventories of nominals involved. We start from the Smith-Stark Hierarchy and extend this approach to additional numbers (such as dual and paucal). Associative plurals appear to undermine this typology, if we treat them as a third number. Either the associative plural or the ordinary plural proves to be exceptional.

The emergence & development over time of alternating stem suppletion among Russian nouns is investigated in a corpus of texts & etymological dictionaries, with particular attention to the inflectional paradigms of two nouns, god 'year' & chelovek 'person'. Stem suppletion is viewed as an intermediate diachronic stage in which one stem is encroaching on the paradigm of another, leading to the full replacement of the latter by the former. Suppletion is found to be conditioned by four principal factors: (1) a paradigmatic leveling tendency, (2) an anomalously high absolute frequency of the lexeme, (3) an anomalously high relative frequency of the paradigmatic slot where suppletion occurs, & (4) the historical semantic reinforcement of one of the suppletive forms in association with particular syntactic constructions. 16 Tables, 16 References. J. Hitchcock

A category often held to be prototypically inflectional, namely number, proves less uniform cross-linguistically in its inflectional status than was once thought (Booij 1993, 1996; Van Marle 1996).2 We therefore examine a series of hypotheses as to its status in section 2. We then take a typological view: number has been the basis for some seemingly robust typological claims and yet these too are problematic (section 3). They can be rescued by clarifying the domain of morphology to which the different claims apply. This leads back to the nature of inflection, with the conclusion that the ‘obligatoriness’ criterion for inflection requires greater prominence than in some recent accounts. (section 4). This criterion also allows us to make progress in understanding facultative number (section 5) and minor numbers (section 6).

In this concluding chapter we draw out and develop some general themes which have emerged in our study of gender; we will also take a look backwards at earlier work and forwards to what may be achieved by future research. The notions of meaning and form provide an entry point for reviewing parts of our study and some previous research (section 10.1). The review of earlier work leads us to a discussion of the development and loss of gender systems (section 10.2). Finally, we look at the prospects in this area, both for understanding major questions of the function of gender and for feasible shorter-term projects (section 10.3).Meaning and formThe relationship between meaning and form is central to linguistics and, not surprisingly, the theme runs through our investigation of gender (section 10.1.1), and through earlier work on the subject (section 10.1.2).A perspective on gender systemsWe saw in chapter 6 how establishing the existence of a gender system and determining the number of genders requires evidence from agreement (that is, evidence concerned with form). At the same time, gender always has a semantic core: there are no gender systems in which the genders are purely formal categories. As shown in chapters 2 and 3, nouns are assigned to gender according to semantic and formal criteria. At one end of the range we find languages like Tamil, in which the meaning of a noun is sufficient to assign it to a gender.

A defective word is defined by paradigm as incomplete compared with the major class it belongs to. Defectiveness signifies the unwanted intrusion of morphological idiosyncrasy into syntax. Although this phenomenon has been a constant subject of studies, it has been ill incorporated into the theories of language. This present volume brings together scholars from various theoretical schools for an overdue typological view of defectiveness. It concentrates on some samples of idiosyncratic gaps which are assumed as indicative of the phenomenon of defectiveness. Before delving into the specified topics of each chapter, this introductory chapter presents a typology of defective paradigms. It discusses terms used to describe defectiveness in synchronotic terms, and the possible diachrony of defective paradigms.

We report a study of the acquisition of colour terms by speakers of Setswana, the language of Botswana in Southern Africa. This was carried out as a test of Berlin & Kay's theory of colour term universals, on a language with less than the maximum complement of eleven basic colour terms, and in order to document changes in Setswana under the impact of economic development. Seventy-seven five- to nine-year-olds were studied on two colour tasks: elicited lists and colour naming. In general the data were consistent with Berlin & Kay's theory: the rank order of frequency of correct use of colour terms was similar to the order of the Berlin & Kay hierarchy; and primary colour terms were offered more frequently and were more likely to be used correctly than secondary colour terms. The use of English colour terms was prevalent, especially amongst the younger groups, but they functioned as substitutes for Setswana terms, rather than as a means to fill the vacant basic colour term slots.

We will review briefly the route we have taken, and then draw out some threads which have run through the book, mainly in order to consider future prospects. We started by identifying all the number values we could find, whichever type of nominal showed them (chapter 2). Then we kept the values still and looked at the different types of nominal involved in the simplest singular–plural systems. We established that their distributions were constrained by the Animacy Hierarchy (chapter 3). Next we put the possible systems of number values together with the hierarchy and found that this typology covered a great deal of the data and the variation, but had to be elaborated to account for phenomena such as conflated number (chapter 4). Then we turned to the means of expression of number, and again found great diversity (chapter 5). The initially daunting difficulties of agreement in number were made manageable, once we drew a clear distinction between controller and target number (chapter 6). In chapter 7 we examined the ‘other’ uses of number, which are many and varied, and found that they too begin to fit into a typology. And then in chapter 8 we turned to verbal number, showing how different it is from nominal number and examining why.While we have been able to make considerable progress, it is clear that there are many aspects still to be better understood.

We have considered the meanings regularly associated with the different values (plural, dual, paucal and so on). We now come to other uses of number, that is, instances where the regular expression of number is taken over for purposes other than its normal meaning. For instance, in honorific usage, plural forms are often used of a single addressee to indicate respect. The semantic and pragmatic effects of number in such uses cannot be derived in the normal way from the usual meanings of the number values. Given that number is often cited as a straightforward grammatical category, apparently reflecting semantics in a regular way, these other uses are found surprisingly frequently. They occur even in familiar languages: we shall see cases where a particular use identified in some distant language turns out to be rather frequent closer to home.There are three broad groups of these other uses: first there are honorific uses (§7.1), as just mentioned; second there are unexpected uses in the general area of conjoining (§7.2); and finally there are various special uses, affective ones in the main (§7.3). In analysing these uses, there are three questions which will recur. The first is whether all number values are available for the particular use; often there are restrictions, and for affective use it is never the case that all values are available. The second is why these uses can be available, particularly since number frequently is a relatively clear reflection of semantics.

Kay and McDaniel (1978) proposed that Berlin and Kay's (1969) linguistic color universals were based on universal perceptual physiology. If this is so then speakers of languages with relatively few basic color terms should have perceptual structures corresponding to the "missing" linguistic categories - the nascent categories hypothesis. Further, speakers of languages with the full set of terms for the 11 universal categories should still retain the perceptual structure that yields the evolutionary path embodied on the hierarchy - the recapitulation hypothesis. This article tests these complementary hypotheses by comparing speakers of English, Russian, and Setswana - languages that have 11, 12, and 5 basic color terms respectively - using a color-grouping task. The fit to the latest version of the theory (Kay et al. 1991) was closer than to earlier versions of the theory, but even so there were still discrepancies. Thus there is support for the universalist's position (there are strong similarities across languages in color grouping despite linguistic differences), but Berlin and Kay's framework may not encapsulate the full extent of color-category universalism.

Network Morphology is a formally explicit approach to morphology which distributes information across a network in which generalizations can be optimally expressed. Generalizations become available in specific cases by the operation of default inheritance. In this paper we explore the notion of ‘default’ in morphology by means of a Network Morphology analysis of the noun classes and genders of Arapesh — a language which relies on a sophisticated understanding of defaults for a satisfactory treatment (Aronoff, 1992). Our work lends support to Aronoff's account of the Arapesh data. It also reveals a confusion in use of the term ‘default’ by linguists. In one usage of the term, the (‘normal case’) default is that which applies in the absence of blocking information; in the other, the (‘exceptional case’) default is that which applies when some exceptional factors prevent normal processes from applying and necessitate the adoption of some ‘last resort’ solution. Under one reading the default equates with the unmarked case; in the other, it is the marked case which is picked out by the same term.

IntroductionThis is an unusual type of chapter. We are looking at the notion of basicness of color terms by comparing various tests. To do this, we are using languages that are well studied. Often this means the basic color terms in those languages can be taken as given, so that we can then see how well particular types of tests perform in identifying the basic color terms. There are two reasons for doing this. One is that it may tell us something about the different types of tests and therefore about the notion of basicness. The second is a practical point. Given the difficulties of fieldwork in particular places, it is worth looking for tests that are easy to run, quick, and efficient, as opposed to those that are more elaborate. So this chapter is about testing the tests. In addition, it has been found that languages which have a full set of basic color terms may still preserve a color hierarchy. That is to say there is not simply a division between basic and non-basic color terms but rather the terms higher on the hierarchy can still be seen to be “more basic” than those lower on the hierarchy. We shall be looking for tests which can reveal that type of structure.First we outline the Berlin and Kay hierarchy. Then we look at the types of measure available and the statistical techniques for establishing how well they fit with the predictions derived from Berlin and Kay.

In an earlier paper we introduced Network Morphology, an approach to inflectional morphology which relies heavily on the notion of default inheritance (Corbett & Fraser 1993), and showed how a complex set of data pertaining to the inflectional morphology of Russian nominals could be handled in this framework. We extend that analysis here to include a range of phenomena which, we believe, has not previously been described in a single formally explicit framework. The key concept underlying our analysis is that of default inheritance. We introduce this below, together with a lexical knowledge representation language called DATR, which we use to make our analysis fully explicit.2 In Section 3 we describe our approach to gender assignment in Russian; animacy assignment is dealt with in Section 4; and declensional class assignment in Section 5.

An intriguing question, which interests non-linguists as well as linguists, is the way in which nouns are allotted to different genders. The linguist who wishes to establish the gender of a given noun can use agreement as a test (for details see chapter 6). However, the native speaker of the language must know the gender of a noun in order to produce the correct agreements (the evidence which the linguist uses). The amount of information is substantial, since native speakers know the gender of many thousands of nouns. For foreign learners of the same language, in contrast, this knowledge often proves elusive in the extreme. How then does a native speaker know the gender of a particular noun? One possible answer would be that the speaker simply has to remember the gender of each noun. This suggestion would involve a considerable feat of memory. It seems an unlikely answer, though many linguists have been ready to accept it. For example, in an often quoted remark, Bloomfield (1933: 280) claimed that:There seems to be no practical criterion by which the gender of a noun in German, French, or Latin could be determined.This pessimism now appears misplaced in view of the following evidence. First, native speakers typically make few or no mistakes in the use of gender; if the gender of every noun were remembered individually, we would expect more errors. Second, words borrowed from other languages acquire a gender, which shows that there is a mechanism for assigning and not just remembering gender.

Number is the most underestimated of the grammatical categories. It is deceptively simple, and is much more interesting and varied than most linguists realize. This was recognized by Jespersen: ‘Number might appear to be one of the simplest natural categories, as simple as “two and two are four.” Yet on closer inspection it presents a great many difficulties, both logical and linguistic’ (Jespersen 1924: 188). Lyons too pointed out its interest: ‘The analysis of the category of number in particular languages may be a very complex matter’ (Lyons 1968: 283). This book will illustrate the interest of number, and some first pointers are given in §1.1.We shall also see the challenges which Jespersen and Lyons allude to, one of the trickiest being the need to ensure that as we compare across languages we are really comparing like with like (§1.2). Hence the book is structured so as to work upwards from properties that are safe building blocks for comparison (§1.3). Finally in this introduction a few notes on presentation are needed (§1.4).The special interest of numberDespite the significance of number, there are still surveys of linguistics where it receives a footnote's worth of attention. This is largely because there are some reasonable but incorrect assumptions about number, which are generally based on the consideration of a rather limited range of languages.

Contemporary linguistic theories distinguish the principal element of a phrase - the 'head' - from the subordinate elements it dominates. This pervasive grammatical concept has been used to describe and account for linguistic phenomena ranging from agreement and government to word order universals, but opinions differ widely on its precise definition. A key question is whether the head is not already identified by some other, more basic notion or interacting set of notions in linguistics. Heads in Grammatical Theory is the first book devoted to the subject. Providing a clear view of current research on heads, some of the foremost linguists in the field tackle the problems set by the assumptions of particular grammatical theories and offer insights which have relevance across theories. Questions considered include whether there is a theory-neutral definition of head, whether heads have cognitive reality, how to identify the head of a phrase, and whether there are any universal correlations between headedness and deletability.

Features allow us to capture regularities in different components of the grammar. Some features operate just within a single component. For example, phonological features are solely a part of phonology: only rules of phonology may refer to them. But there are also features which apply across component boundaries, such as the well-known instance of morphosyntactic features (Matthews 1972: 162). The distinction has been around for many years, but Svenonius (2003: 376) suggests helpful terms: features available within a single component are ‘internal’ features, while those which apply across components are ‘interface’ features.Provided we can define and distinguish the feature types clearly, we can maintain some interesting claims. First, syntax is phonology-free (Zwicky & Pullum 1983, Pullum & Zwicky 1988); that is, syntactic rules cannot refer to phonological features. A rule of the type ‘vowel-initial verbs take clause-final position’ is excluded. And second, syntax is also morphology-free (Zwicky 1992a: 354–56, Corbett & Baerman 2006; see §3.4.8 below); that is, syntactic rules cannot refer to morphological features. This principle excludes rules of the type ‘verbs of inflectional class ii take clause-final position’.

We began by trying to write a grammatical description without using features. The account was repetitious, and missed the point. The attempt illustrated vividly why features are central in mainstream linguistics. They have a key place for good reason, and they have important roles in linguistic frameworks which are very different in other respects. But now consider what languages would be like if we genuinely did not need features. If for instance the sound systems of natural languages were not naturally described in terms of a relatively small number of oppositions. Or if there were no features like number, running through the morphosyntax and affecting items of different parts of speech. We would lose much of the intriguing richness of natural languages. If we look at inventories, the five-valued number system of Sursurunga, with its paucal and greater paucal, is impressive. The simple systems of English or Russian are not so obviously challenging, until we look at the nouns involved in the number system and find that the border for number-differentiation is drawn at slightly different points. case in German seems trivial compared with the massive case system of Archi, until we look at the tricky patterns of syncretism in the German system.If we turn to the way in which features and values are distributed, some languages like Russian look easy at first: verbs mark the person, number and gender of the subject. We then compare with languages like Archi where the rules of agreement are more complex, and moreover many verbs do not agree. There is therefore plenty to be done, for those who care about what the features mean, and for those who are concerned with their distribution, from fieldworkers and typologists through to the computational linguists. More importantly, there are strong reasons for bringing these perspectives together. We need to make reasoned rather than habit-based choices about our use of features as we analyse and enjoy the richness and diversity of the world's languages.

Gender resolution is an area in which the data are often surprising and interesting, yet the topic is frequently left out of account. The term ‘resolution rule’ is taken from Givón (1970), and it refers to a rule which specifies the form of an agreeing element (or target) when the controller consists of conjoined noun phrases. If we have a sentence like Mary and John are happy, it is the number resolution rule which specifies the use of the plural are, rather than is. If we translate this sentence into a language like French, where predicative adjectives agree in gender, we need a gender resolution rule to establish the gender of the adjective, since one conjunct is feminine and the other is masculine. We shall also meet more complex cases: in Slovene, if a neuter singular and a feminine singular are conjoined, it is the gender and number resolution rules which specify the form of the target, say the verbal predicate, as masculine dual. This example, like the English one, illustrates the point that resolution rules do not operate only to resolve feature clashes but can also operate when conjuncts share features (singular in this example). It also suggests that this topic draws together problems connected with controller genders and target genders. While gender resolution will be our main concern, we should see it in the wider context of feature resolution (section 9.1).

Imenice pluralia tantum su fascinantna pojava, za koju se slavisti interesuju još od Braunove teze (1930), pa i ranije. Za to postoji nekoliko razloga. Prvo, mnoge od ovih imenica su defektne jer, iako su brojive, nemaju jedninu. Drugo, dok defektne imenice obično uključuju sporadične praznine u promeni (kao što je to, po jezičkom osećanju mnogih govornika, slučaj sa genitivom množine ruske imenice *mečt), klase imenica pluralia tantum često su – bar delimično – semantički predvidljive (tj. ove imenice su podložne „uopštavanjima srednje veličine”, Koenig 1999). Međutim, ova predvidljivost je ograničena: up. u ruskom imenicu binokl ‘dvogled’, koja ima standardnu promenu, sa imenicom sani ‘sanke’, koja je imenica plurale tantum. Treće, da bi se mogle upotrebiti kao brojive, imenice pluralia tantum često zahtevaju neko dodatno sredstvo, npr. zbirne brojeve. I u opštoj lingvistici postoji stalno interesovanje za ove imenice (npr. Visnievski 2009). Zato je glavni cilj ovog referata da pokaže poseban značaj podataka iz slovenskih jezika za opštelingvističko ispitivanje imenica pluralia tantum. Taj značaj je dvostruk. Prvo, kod nekih klasa ovih imenica se uočavaju zanimljive razlike u različitim slovenskim jezicima. I drugo, treba preispitati tvrdnju da je, iz semantičkih razloga, prirodno da se predmeti koji se javljaju u paru označavaju imenicama pluralia tantum, s obzirom da se ovakvi predmeti označavaju imenicama pluralia tantum i u jezicima koji imaju dual (kao u slovenačkom i gornjem i donjem lužičkom). Uz to, u ovim jezicima se imenice pluralia tantum mogu koristiti za jedan, dva ili više od dva referenta (kao u donjem lužičkom, Janaš 1976/1984). Dakle, ove imenice zaslužuju našu stalnu pažnju bilo da je reč o razumevanju broja u slovenskim jezicima ili o opštelingvističkim aspektima imenica pluralia tantum. = = = Pluralia tantum nouns are indeed fascinating, and Slavists’ interest in them goes back to Braun’s thesis (1930) and earlier. There are several reasons for this. First, many of these nouns are defective, since they are countable yet they lack a singular. Second, while defectives typically involve sporadic gaps (as with Russian genitive plural *mečt for many speakers), sets of pluralia tantum nouns are often semantically predictable, at least in part (they are subject to ‘middle-size generalizations’, Koenig 1999). This predictability is limited, however: compare Russian binokl ‘binoculars’ (which is a normal noun) with sani ‘sleigh’ a plurale tantum noun. Third they often require some sort of repair (such as the use of collective numerals) so that they can be treated as count nouns. Pluralia tantum nouns are of continued interest in the general linguistic literature (see, for instance, Wisniewski 2009), hence it is timely to consider the particular interest and contribution of the Slavonic data, the focus of this paper. Two aspects stand out. First the Slavonic languages provide semantic classes of these nouns, which vary across the family in interesting ways. And second, the claim for the semantic naturalness of paired objects being pluralia tantum needs reassessing, given that these nouns are pluralia tantum even when the dual is available (as in Slovene and Upper and Lower Sorbian. Moreover, in these languages pluralia tantum nouns can be used for one, two, or more than two referents (as in Lower Sorbian, Janaš 1976/1984). Thus these nouns deserve our continued attention, both for Slavonic internal and for general linguistic reasons.

Newly elicited data from Mohawk, an Iroquoian language still spoken in six communities of New York, Quebec, & Ontario, are analyzed in support of a claim that all forms of noun incorporation in Mohawk constitute a process of word formation. Mohawk noun incorporation generally takes the form of an intransitive or transitive verb in which a noun stem denoting a backgrounded semantic patient immediately precedes the verb root; person-marking prefixes apply to the resulting lexical compound & never corefer to the incorporated item, which therefore is neither a syntactic argument nor in apposition to a pronominal argument. Mark Baker's distinction between lexical & syntactic types of incorporation in Mohawk is shown not to have a sound empirical basis, as (1) syntactic tests used by Baker necessarily involve animate incorporation, which is marginal in Mohawk, & (2) his external argument alternatives differ significantly in information structure from incorporation. Morpheme-specific variation in the productivity, acceptability, & grammatical & semantic transparency of incorporations provides further support for their lexical status in Mohawk. 24 References. J. Hitchcock

When we analyse feature systems in detail, particularly when using the approach described in §4.1.2, we find some surprises. We discover systems whose main outline is clear, yet which also have some remarkable characteristics. Several of these have been catalogued and labelled. Typically these interesting phenomena are described feature by feature. If we take a more abstract view, we can discern regular patterns in the apparent oddities. I therefore suggest a canonical approach (§6.1) and lay out a canonical scheme for morphosyntactic features, using morphological criteria first (§6.2). This allows us to calibrate the phenomena which lie outside the core of feature systems. We find that indeed the same types of non-canonical behaviour are found with different features, though they may be given rather different names in the literature (§6.3). We evaluate these results in §6.4. Then we consider the syntactic principles for canonical features in §6.5 and review canonical inflection in §6.6.The canonical approach in typologyImagine we found a language in which every last noun had robust morphology distinguishing singular and plural, and every verb, adjective and adposition showed clear agreement in number. We would propose a morphosyntactic feature number, with the values singular and plural, without hesitation. Any alternative would make the syntax highly redundant (rather like what we saw in §1.1.1). On the other hand, if we found a language in which the only trace of number was a distinction between the equivalent of ‘I’ and ‘we’, then an argument for a number feature would need skilful support. Of course, many languages fall between these two extremes. Yet we are perhaps too ready to treat them as though they were instances of the first type. Morphosyntactic features, like number, often have a ‘penumbra’ where the data are not clear-cut, and we need to be careful in our analysis.

Language has such a central place in our lives and in research that it is difficult to find an outside vantage point from which to achieve real understanding. As linguists, we attempt to do this by treating language as our object, while restricting the use of language as the tool. As a result of this approach, linguistics is in an exciting phase. Theories compete for overlapping segments of the research space. There is a sense of great achievement in some areas and equally of uncertainty about common goals. In this rapidly changing scene, one constant is the use of features. Fieldworkers, sociolinguists, computational linguists, syntacticians, working on spoken or on signed languages, all standardly use features. They are the key underpinning for linguistic description. We use features a good deal, but sometimes we take them for granted, assuming we all share the same conventions. In reality, the use of superficially similar notations sometimes hides differences in the underlying logic as well as in the substantive semantics of features.It is therefore worth working through the motivation for using features, and the choices available to us. Naturally, different researchers make different choices; the important thing is that these should be reasoned choices, and that they should be made explicit. We shall give special attention to syntax and morphology, since it is in these components that the use of features requires the clearest argumentation. This is because these features do not have direct correspondences in meaning or sound, we have not such immediate evidence for them, and hence must justify their use with particular care. Having isolated the distinctions which we model using features, it is natural to typologize across them. As with all typology, we need to consider carefully whether we are comparing like with like, an issue which we discuss in §5.1.

Contemporary linguistic theories distinguish the principal element of a phrase - the ‘head’ - from the subordinate elements it dominates. This pervasive grammatical concept has been used to describe and account for linguistic phenomena ranging from agreement and government to word order universals, but opinions differ widely on its precise definition. A key question is whether the head is not already identified by some other, more basic notion or interacting set of notions in linguistics. Heads in Grammatical Theory is the first book devoted to the subject. Providing a clear view of current research on heads, some of the foremost linguists in the field tackle the problems set by the assumptions of particular grammatical theories and offer insights which have relevance across theories. Questions considered include whether there is a theory-neutral definition of head, whether heads have cognitive reality, how to identify the head of a phrase, and whether there are any universal correlations between headedness and deletability.

In this paper we introduce a declarative approach to inflectional morphology, which we call Network Morphology, using the lexical representation language DATR. We show that we can account for a range of (Russian) data, for which previously various rule types were required, and can provide a more satisfying analysis than was previously available. First we outline the essential data (section 2), highlighting the problems they present. Section 3 introduces the basic tenets of Network Morphology. This draws heavily on DATR, which we present in outline in section 4. Next we reconsider the Russian declensional classes from this new perspective (section 5). We show how the approach described overcomes long-standing problems in an elegant fashion; the complexity of the data suggests that the approach adopted has implications well beyond Russian. We then tackle the complex problem of animacy in Russian, which exemplifies interesting regularities extending across declensional classes (section 6).

We report a study of the colour terms of Setswana, involving 390 subjects, which compares children with adults, and people from rural areas with people from more urban areas. The study was carried out as a further test of Berlin & Kay's (1969) theory of colour universals and as an investigation of the variations and developments in Setswana across age groups and area of residence. Our results show a move away from traditional Setswana colour terms towards the use of borrowed English terms across the entire colour term inventory, particularly in the young and those who have been to school. Further, a decline in the use of Setswana ‘cattle’ terms by the young reflects a change in tradition. The basic colour term inventories were by and large consistent with Berlin & Kay, irrespective of whether these were borrowed English terms or traditional Setswana terms. Finally, an unusually large number of the sample showed partial colour vision defects.

Recall why we use features: we wish to capture generalizations across linguistic objects. Given the items children, books, saxophones, we want to state that they are all plural, and take plural agreement. We could use plural as a label, and go on to give each object a set of such labels. But the key point is that plural is interesting because of singular. Child is not just not plural, it is singular. Thus plural is more than a label to be attached; it signals something of linguistic interest, from which other information can be derived. It is a member of an opposition, that is, it is a value of a feature, in this instance number. In a language like English, the speaker has to choose singular or plural. The choice is obligatory (an issue to which we return in §2.1.6 and in §6.5); we meet a very different type of number system in §8.1.Various notations have been employed, including these: children [+plural] or children [+pl]children [plural +]children [plural 1]children [plural=1]children: =1children: =yesplural (children) = trueThe ± notation is often used with binary (sometimes called Boolean) features, as in Gazdar, Klein, Pullum & Sag (1985: 22–3); see §2.1.2 below. The notations given do not reflect significant differences. I will suggest an appropriate formalism, making the feature-value structure clear, in §2.1.2. We shall look at the general issue of the structuring of features (§2.1). Then we consider a set of specific issues relating to structuring: atomic versus complex values (§2.2), the way in which features cross-classify (§2.3), typing (§2.4) and unification (§2.5).

There are fascinating problems at the syntax-morphology interface which tend to be missed. I offer a brief explanation of why that may be happening, then give a Canonical Typology perspective, which brings these problems to the fore. I give examples showing that the phenomena could in principle be treated either by syntactic rules (but these would be complex) or within morphology (but this would involve redundancy). Thus ˋnon-autonomous' case values, those which have no unique form but are realized by patterns of syncretism, could be handled by a rule of syntax (one with access to other features, such as number) or by morphology (with resulting systematic syncretisms). I concentrate on one of the most striking sets of data, the issue of prepositional government in Latvian, and outline a solution within Network Morphology using structured case values.

We have looked at the evidence we use to justify postulating particular features and their values in particular languages. That evidence, especially from agreement and government, is what is available to the child learner and to the linguist. Given the genuine difficulties of analysis we have found, we might wonder what more general claims we could make when we turn to comparison across languages. A simple suggestion (Zwicky 1986a: 988) is that ‘universal grammar should permit only a finite number of attributes and values – indeed […] universal grammar should provide finite lists of the attributes and values available for service in a particular grammar’. This idea has since been discussed within Minimalism (thus Chomsky 2001: 10 appears to take a similar view). Zwicky points out the difficulty with the approach, as put to him by Gerald Gazdar: He [Gazdar] observes that there is a serious correspondence problem involved in talking about ‘the illative case’ in two different languages: what allows us to identify the two grammatical cases? Similarly for other agreement properties […]I [Zwicky] believe it is possible to require that every property on the lists have semantic concomitants. I am not maintaining here that these properties are to be identified with semantic features; grammatical categories are virtually always arbitrarily distributed (from the semantic point of view) in the lexicon to some extent.Zwicky (1986a: 988–9)Zwicky's suggestion, then, is that morphosyntactic features (‘properties’ in his terms) always have a semantic core, and it is this core which allows comparison across languages. They are often partly arbitrary, but never fully arbitrary. We return to this issue in §5.1.

IntroductionA major focus of the debate on headedness has been the problem of determining the head in different constructions and of establishing acceptable criteria to enable us to do so. The data have been taken mainly from English, and so this account, by contrast, extends the investigation to a language with a much richer morphological system than that of English, namely Russian. We shall concentrate on numeral expressions in Russian, where the head-dependent relation has long been known to be problematic (see, for example, Isačenko, 1962: 529). We shall examine them in the light of the criteria for heads proposed by Zwicky (1985) and by Hudson (1987). At first sight it seems that no single head can be identified for these constructions; rather, the properties of the head appear to be shared between different elements, which would fit with Zwicky's approach. However, given current assumptions about lexical entries and feature distribution, these constructions can be analysed as being rather less exotic than they first appear, and as having a consistent head, as Hudson would predict. While attempting to remain as theory-neutral as possible, we shall develop the analysis to see whether the idea of a single element having all the head properties can be maintained. It is in focusing on the question of headedness that this chapter differs from most previous accounts of Russian numeral phrases. We shall see that there are two consequences. The first is that we still need to recognize that headedness is a gradient notion: a particular element may have head-like characteristics to a greater or lesser degree, and that these may vary according to external factors (notably, case assignment).

As with any theoretical construct, we should justify every feature and value that we use. We do this both for the account of each specific language, and at the level of our general theory. We shall aim for a list of the features and their values. This would be a beautifully simple typology. It therefore makes good sense to work towards such an inventory, unless and until it is proved impossible. Establishing this inventory requires the solution to two problems, the analysis problem and the correspondence problem. In this chapter we concentrate on the analysis problem – how we determine the features and their values in a given language. In the next, we go on to the correspondence problem – whether the features and values we identify are in some sense the same across languages. Our main focus in the current chapter, then, is the analysis problem, the justification of features and their values. As we make a first attempt at the analysis problem (§4.1), three issues arise and need to be dealt with for us to make further progress. First the question of conditions (§4.2), then the role of hierarchies (§4.3), and third, the serious problem of gradience (§4.4). Having tackled these issues, we consider canonicity briefly (§4.5).The analysis problemWhen tackling the analysis problem we face a whole set of questions. How do we decide if a language has a particular feature? For example, does Archi have person? How do we decide how many values a feature has in a given language? For example, how many case values does Russian have? Some of the issues are profound, and will be the subject of continuing debate, while others are highly practical, concerning standardization and the presentation and glossing of examples. We begin with the analysis problem at the level of features (§4.1.1) and we discuss values shortly, in §4.1.2; we give special attention to gender values in §4.1.3, and consider the question of an inventory in §4.1.4.

The Surrey Syncretism Database encodes information on inflectional syncretism in 30 genetically and geographically diverse languages, representing such morphosyntactic features as case, person, number and gender. Syncretism is defined as when some set of words fail to distinguish morphosyntactic feature values which we believe, based on language-internal criteria, to be underlyingly present (for example, in Latin, the dative and ablative cases may be distinct in some contexts but collapsed into a single form in others). For each language all instances of syncretism are recorded.

It is now time to look at the ways in which number is expressed. This rather basic issue is a surprisingly novel one. Of course, many grammars describe number marking in individual languages but little has been done towards a typology. We shall therefore give an initial typology, one which aims to list the possibilities (claiming that the listed types can exist and no more). The obvious candidates for number expression are all found: special words (§5.1), syntax (§5.2), morphology (§5.3) and lexical means (§5.4). We shall then examine three types of system which are distinctive and which belong here in a discussion of means because they do not give rise to new semantic distinctions. (This is why they were not treated in chapter 2.) These are inverse systems (§5.5), minimal-augmented systems (§5.6) and ‘constructed’ numbers (§5.7). Finally we take up the discussion of the reduced expression of number, considering the form of items which are not (or not fully) within the number system (§5.8).The main typological point is the importance of comparing like with like. In previous chapters we examined number values and their ranges of availability and compared each of these across languages. Now we turn to the means of expression, and must continue to be clear about when our claims relate to meaning and when to the means of expression.

Imagine that it mattered to us how many teeth a giraffe has, or how many stomachs a llama has. Perhaps unaware of the potential difficulties, we might expect zoologists to be able to give us answers to such queries. Linguists have been asked comparable questions increasingly frequently in recent years, as various researchers attempt large-scale comparisons, based on seemingly straightforward measures like the size of the phoneme inventory, the number of morphosyntactic feature values, and so on. Yet the question ‘how many case values has Russian?’ was mentioned earlier (§4.1.2, footnote 4), as a difficult question that has given rise to a considerable literature. Why then is the task difficult? We expect the zoologists to have rounded up a few healthy giraffes, persuaded them to open their mouths (!), and to have counted and recounted the number of teeth. The linguistic question requires instead the specification of a system. Rounding up a few Russian nouns will not do. We want an account of all the possibilities, which means checking other parts of speech too and different contexts. But having established the largest number of possibilities we also want a realistic picture of how the case values are distributed across the lexicon. While we might be forgiven for not knowing the issues of classification and definition which zoologists face, we should be aware of the dangers of simple answers to apparently simple questions within our own discipline.We have seen already (§§4.1.2–4.1.3) that establishing feature values can be challenging. We now look at a single language and find that the difficulties we identified scattered across different languages can co-occur in one. We analyse case in Russian, partly because we have so far given less attention to case, partly because there is a tradition of debating the case values of Russian, which was significant in the development of the Set-theoretical School. But mainly, we look at case in Russian to demonstrate that features are not ‘clean and neat’, nor even clean and neat apart from occasional exceptions. They are more interesting than that. Russian certainly presents a worthy challenge, with different analyses giving it as few as six case values or possibly as many as eleven. Given the preparation in §4.1.2 we can now take on this challenge. Furthermore, the canonical approach, which uses a logical scheme against which to evaluate the different values proposed (§6.2), will prove equally useful. We go into some detail, partly to prove the points claimed, but also to show the complexity of a feature system when analysed thoroughly.

IntroductionIt is a generally shared assumption that any adequate formal model of morphology ought to take some account of inflectional syncretism. Most investigators who have addressed syncretism overtly have taken this assumption one step further: it is not enough to describe syncretism, one should also constrain it. This goal is motivated by two factors. First, syncretism is something of an aberration: by default we assume a one-to-one relationship between morphosyntactic function and form, and syncretism is a violation of this assumption (Carstairs 1987). Second, syncretism displays preferred patterns, as we have seen throughout Chapter 3.Ultimately, the morphological description of a particular syncretism must contain two elements: (i) a list of the set of values which are syncretic, and (ii) a way of associating this set with a form. To a large extent, constraints on syncretism are a product of how these elements are treated. For example, the syncretic set may be a natural class of values or simply a stipulated disjunction, while the form itself may be defined over the whole set, or defined in terms of one of the constituent members. In §4.2 below we examine the inherent properties of different rule types, and in §4.3 we see how these have been employed in particular accounts of syncretism.Defining sets of valuesNatural classesProbably the most common approach to syncretic sets of values is to treat them as a reflection of underlying feature structure.

Con bibliografia e indice. Morphological typology of deponency / Matthew Baerman -- Deponency, syncretism, and what lies between / Greville G. Corbett -- Extending deponency: implications for morphological mismatches / Andrew Spencer -- A non-canonical pattern of deponency and its implications / Gregory T. Stump -- Deponency in the diachrony of Greek / Nikolaos Lavidas and Dimitra Papangeli -- Deponency in Latin / Zheng Xu, Mark Aronoff, and Frank Anshen -- Declarative deponency: a network morphology account of morphological mismatches / Andrew Hippisley -- The limits of deponency: a network morphology account of morphological mismatches / Andrew Hippisley -- The limits of deponency: a Chukoto-centric perspective / Jonathan David Bobaljik -- Slouching towards deponency: a family of mismatches in the Bantu verb stem / Jeffrey Good -- Spanish pseudoplurals: phonological cues in the acquisition of a syntax-morphology mismatch / Ricardo Bermúdez-Otero -- Pseudo-argument affixes in Iwaidja and Hgar: a case of deponent subject and object agreement / Nicholas Evans -- How safe are our analyses? / P.H. Matthews.

In the first chapters of the book it was largely taken for granted that the number of genders in a given language could be readily established, and we tackled the problem of how nouns are assigned to particular genders. Then in the preceding chapter we considered the syntactic means by which gender is manifested. And we saw that in some instances the evidence is complex. We need, therefore, to work out how we determine the number of genders in a given language. In several of the more familiar languages, the gender pattern is straightforward and the way in which the system is analysed is taken as self-evident. In other languages, linguists may present the pattern as though it were equally uncontroversial, but we find that similar situations are described differently by those working on different language families. In contrast, the number of genders in a particular language can be the subject of interminable dispute. Given this unsatisfactory situation, we must develop a consistent approach to analysing gender. For those most familiar with languages which have relatively transparent gender systems, this chapter may appear unduly detailed. But in order to make meaningful comparisons between such languages and those with more complex gender systems we need to ensure that the starting point is the same. After a brief discussion of terminology (section 6.1), we move on to the central notion of ‘agreement class’ (section 6.2). In section 6.3 we investigate how the nouns in an agreement class may make up a ‘controller gender’.

This paper analyses constraints on inflectional syncretism and inflectional allomorphy using frequency information. Syncretism arises where one form is associated with more than one function, whereas inflectional allomorphy occurs where there is more than one inflectional class, and a single function is associated with two or more forms. If high frequency is associated with more differentiation on both sides, we expect, on the one hand, that a frequent function will have a high number of forms and, on the other, that a frequent form will have a high number of functions. Our study focuses on Russian nominals, in particular nouns, which exhibit both syncretism and inflectional allomorphy. We find that there is a relationship between frequency and differentiation, but that it is not exceptionless, and that the exceptions can be understood in terms of the use of referrals as default rules.

We report a study of the acquisition of colour terms by Russian children which had two main aims: first, to test Berlin & Kay's (1969) theory of colour universals using acquisition order as a measure of basicness; and secondly, to see if the two blue terms of Russian are genuinely basic. Two hundred children aged from three to six-years-old were tested on three colour-tasks – colour term listing, colour term production and colour term comprehension. To a reasonable approximation, the order of colour term acquisition was in accord with Berlin & Kay's theory, but the data are also consistent with the weaker claim that primary terms tend to be learned before derived terms. On balance the data were consistent with Russian exceptionally, having an extra term for the blue region. But, the two blue terms – goluboj ‘light blue’ and sinij ‘dark blue’ – were confused more often than other pairs of terms even by the five- to six-year-old sample.

This paper has three goals. First we wish to elucidate the complex paradigms of Dalabon. In languages like Dalabon, which encode information about multiple pronominal arguments on adjacent slots on the verb, the two slots are frequently merged. The resultant set of combinations typically lies somewhere between an irregular paradigm and a set of forms derivable by combining subject and object elements according to some set of rules. These paradigms are potentially vast — in Dalabon, which has a rich set of person, number and kinship categories in its pronoun system, there are 102 possible subject/object combinations, each of which further distinguishes six tense/aspect/mood categories. Most languages of this type reduce the number of forms by widespread identities of form within the paradigm. However, it is not always clear whether the formal collapse is accidental homophony or principled syncretism.

To investigate the relationship between morphological irregularity & word frequency, plural nouns in the one-million-word Uppsala corpus of Russian are categorized for three types of frequency anomalies & seven degrees of irregularity. For each degree of irregularity, probability values are calculated for the absolute frequency of plural forms, their relative frequency compared to singular forms of the same lexeme, & the frequencies of paradigmatic case cells relative to those of the same cells throughout the corpus. Results strongly confirm a hypothesis that absolute plural anomaly correlates with irregularity; ie, nouns with irregular plural subparadigms tend to occur frequently in the plural. Relative plural anomaly was found to correlate significantly with segmental irregularity only, not stress irregularity; no relationship was found between irregularity & cell frequency. 7 Tables, 2 Figures, 1 Appendix, 27 References. J. Hitchcock

Number is the most underestimated of the grammatical categories. It is deceptively simple yet the number system which philosophers, logicians and many linguists take as the norm - namely the distinction between singular and plural (as in cat versus cats) - is only one of a wide range of possibilities to be found in languages around the world. Some languages, for instance, make more distinctions than English, having three, four or even five different values. Adopting a wide-ranging perspective, Greville Corbett draws on examples from many languages to analyse the possible systems of number. He reveals that the means for signalling number are remarkably varied and are put to a surprising range of special additional uses. By surveying some of the riches of the world’s linguistic resources this book makes a major contribution to the typology of categories and demonstrates that languages are much more varied than is generally recognised.

In chapter 2 we investigated number values, finding up to five in certain languages. While we were trying to find all the possible systems, we looked at whichever type of nominal, whether pronoun or noun, was most promising in a given language. Then in chapter 3 we asked which nominals can be part of the number system, and to do so we concentrated on the basic singular–plural opposition. Ideally we would like to integrate these two dimensions of variation, to achieve a typology which predicts the possible patterns of values available at different points on the Animacy Hierarchy. With this goal, we begin with the strongest hypothesis: the Animacy Hierarchy constrains the distribution of all number values. We shall see that this claim needs some elaboration but that, perhaps surprisingly, it is basically true (§4.1). We examine potential counter-examples: minor numbers (§4.2), associatives (§4.3) and distributives and collectives (§4.4). Then we examine languages with different systems at different points on the hierarchy (§4.5).The chapter illustrates an important approach in typology and the type of challenges which arise. Having established the basics, that is, the number values which languages may have and the patterns of values available for different nominals, we now combine those two elements to construct a complete typology. At several points we find that we ‘run out’ of languages: the critical combination of factors may be rare or non-existent (or has not yet been found).

An attempt to determine some of the requirements of that portion of linguistic theory concerned with agreement. Within Slavonic there are numerous difficulties, which may be classified into three types, according to whether they are caused by the controller, the target, or the agreement features. There are difficulties in identifying the controller, in establishing its features, & in handling reluctant controllers (which do not appear in surface structure in a form that matches that of the target). The major target problems involve compound targets & reluctant targets. Features require resolution rules & rules to assign default values. Various constructions discussed permit agreement choices, whose distribution is determined by a range of linguistic & sociolinguistic factors. Patterns can be established for the use of agreement options, but some of these must be stated at corpus level rather than at sentence level. The conclusion is that even within a closely related & relatively conservative set of IE langs there are several constructions that are problematic for linguistic theory. 2 Tables, 43 References. HA

An attempt is made to establish the basic/nonbasic status of Russian color terms in the context of B. Berlin's & P. Kay's color hierarchy (Basic Color Terms: Their Universality and Evolution, Baltimore, 1969), using W. F. Battig's & W. F. Montague's method of psychological salience ("Category Norms for Verbal Items in 56 Categories: A Replication and Extension of the Connecticut Category Norms," Journal of Experimental Psychology, 1969, 80, 3, 2). Russian native speakers (N = 31) were asked to list as many color terms as they could think of in five minutes. After each minute, they were asked to draw a line on their paper before continuing. Color term ranking is tabulated on the basis of first, second, third, fourth, & fifth minute periods & the number of responses for each term after successive minutes. The concentration of indisputably basic color terms at the top of the table was attested & the terms siniy 'dark blue' & goluboy 'light blue' were listed as basic terms. 1 Table, 8 References. Z. Dubiel

This text presents a critical overview of current work on linguistic features - gender, number, case, person, etc. - and establishes new bases for their use in the study and understanding of language.

During the ten years that I have been working on this book, many people have offered data, ideas, comments and other help. It is a pleasure to record my thanks to them: Alexandra Aikhenvald, Peter Austin, Matthew Baerman, Julia Barron, Laurie Bauer, Barry Blake, Jim Blevins, Juliette Blevins, Misi Brody, Jurgen Broschart, Dunstan Brown, Gordon Brown,Wayles Browne,Wallace Chafe, Hilary Chappell, Ross Clark, Ulrike Claudi, Richard Coates, Michael Colenso, Stanford Cormack, Alan Cruse, Anna Morpurgo Davies, Alan Dench, Aleksandra Derganc, Ivan Derzhanski, Werner Drossard, Barrie Evans, Nick Evans, Roger Evans, Ray Fabri,William Foley,Michael Fortescue, Victor Friedman, Norman Fraser, Gerald Gazdar, Adele Goldberg, Ian Green, Rebecca Green, Ekaterina Gruzdeva, Tom Guldemann, Ken Hale, Shelly Harrison, Mark Harvey, Martin Haspelmath, Katrina Hayward, Bernd Heine, George Hewitt, Jane Hill, Nikolaus Himmelmann, Andrew Hippisley, Robert Hoberman, Richard Hogg, Axel Holvoet, László Honti, Jim Hurford, Don Hutchisson, Larry Hyman, Ruth Kempson, Aleksandr Kibrik, Simon Kirby, Klaus-Michael Köpcke, Antonina Kovaĺ, Ulla-Maija Kulonen, Klaus Laalo, Mary Laughren, Velma Leeding, Werner Lehfeldt, Frank Lichtenberk, Jouko Lindstedt, Elizabeth Löbel, Paul Marriott, Michael Mitchell, Arto Mustajoki, David Nash, Johanna Nichols, Almerindo Ojeda, Janez Oreŝnik, Barbara Partee, Frans Plank, Maria Polinsky, Tom Priestly, Robert Ratcliffe, Nick Reid, Bruce Rigsby, Jan Rijkoff, John Roberts, John Saeed, Hans-Jurgen Sasse, Wolfgang Schellinger, Harold Schiffman, Jane Simpson, Elena Skribnik, Neil Smith, Gerald Stone, Greg Stump, Roland Sussex, Dalija Tekoriene?, Hannu Tommola, Larry Trask, Robert Van Valin, Nigel Vincent, Max Wheeler, David Wilkins and Dieter Wunderlich.

We saw in the last chapter that agreement is one of the ways of expressing number and this will be our main topic here. We first consider agreement and the types of mismatch which occur between the controlling noun phrase and the agreement target. Then we undertake a set of case studies leading to a typology of agreement options in number. There are three typological themes: once again we see the importance of being clear about terms, as we analyse systems where the number values of controller and target differ; we see the importance of hierarchies, this time in syntax; and there is a graphic illustration of how factors which can be identified as being at work in various languages interact in different ways to give very different results.Controller versus target numberThe first notion that we need is ‘agreement’, which is the covariance or matching of feature specifications between two separate elements, such as subject noun phrase and verb. We shall call the element which determines the agreement (say the subject noun phrase) the controller. The element whose form is determined by agreement is the target. The syntactic environment in which agreement occurs is the domain of agreement. And when we indicate in what respect there is agreement, we are referring to agreement categories or agreement features, as illustrated in figure 6.1. Of course, we are primarily interested in agreement in number. The controller of agreement is a noun phrase. Cross-linguistically, the possible targets are more varied than many believe.

Taking stockIt is worth reflecting on how we have reached this point. We began from a logical typology of syncretism. We asked what might theoretically be found, working up from the simplest possibilities (Chapter 2). This approach is feasible because the phenomenon is finite, and particularly worthwhile because syncretism is complex, and the reality does not correspond to many linguists' hunches about it. Given this basis we could move on to the cross-linguistic typology of Chapter 3, where we saw that extending the range of languages beyond those most usually treated for syncretism changes the picture dramatically. Among other things it shows that some of the patterns of syncretism which are familiar from Indo-European are actually somewhat exotic. We relied largely on two samples. First of all, our own sample for the Surrey Syncretisms Database, with thirty languages chosen for genetic and geographic diversity but with the entry condition that they must show instances of syncretism. These were investigated in great detail in preparation for this book. The second sample is that of the World Atlas of Language Structures. This sample of 200 languages, also chosen to avoid genetic and areal bias, provides a useful balance in that it was externally selected; of course, some of the languages provided no helpful data since they lack inflectional morphology. While the breadth of these samples is invaluable in ensuring coverage, so that unusual phenomena do not fall through the net, we were not tied to them.

We shall map out the range of data to be considered. We include problems familiar from the widely cited languages, but we shall also considerably extend the scope of the discussion and of the languages investigated. We first look at the patterns of syncretism and their implications (Chapter 2: §2.1), and then examine the domains which allow us to compare paradigms (Chapter 2: §2.2). This permits us to begin considering the types of analysis available (to be discussed fully in Chapter 4). In Chapter 2: §2.3 the important issue of directionality is raised. Then we consider the ‘extreme’ interpretations in Chapter 2: §2.4, namely neutralization and uninflectedness. These prove to describe the easy instances: the more challenging ones lie in between and form the subject of the remaining chapters.Syncretic paradigmsTypesAs we have characterized it, syncretism involves the identity of cells within an assumed morphosyntactic paradigm. Graphically, we shall represent this by generating a complete inflectional paradigm, and enclosing the identical forms within a box. Before considering what morphosyntactic values may form the parameters, let us outline a brief typology of the ways paradigmatic cells may be united. These are illustrated below with examples of case syncretism. The different types will turn out to have important consequences for the representation of feature structure.In the simplest pattern, which we call simple syncretism, two or more cells with different values for a feature are merged.

Understanding the complexity of natural language is one of the great intellectual challenges. As linguists we try to do this through a variety of approaches and theories. For all our differences, one thing that most linguists share is the use of features. Features allow us to identify common properties; we propose a feature number, with the values singular and plural, as we find in forms like lake ~ lakes, loaf ~ loaves, woman ~ women. Using a feature like this captures the intuition that lake and lakes are forms of the same word, while on the other hand the plurals (lakes, loaves and women) are also in some sense the same. Though plural number is realized differently on each, they behave identically for agreement, since they all take a plural determiner (these rather than this). Other examples of features include gender (with values such as masculine, feminine…) and person (first, second, third). These are examples of morphosyntactic features. Features may also be semantic, such as animacy (ranging over human, other animate, inanimate), syntactic (for part of speech categories such as verb or noun), morphological (for inflectional class) or phonological (specifying, for instance, the height or backness of a vowel).Features, then, are our means of capturing what is consistent across linguistic entities within a language; they also help us to identify what is consistent across languages. Various languages have a number feature rather similar to that of English, while in others this feature shows interesting differences. Features have proved invaluable for analysis and description, and they have a major role in contemporary linguistics, from the most abstract theorizing to the most applied computational work. As we rely increasingly on features, it is important to review our assumptions and check our progress in understanding them. In particular, there is a tradition in a part of the discipline to be careful about the formal properties of features, being scrupulous about the mechanisms according to which they work within given theories. Another set of linguists have worked hard to understand the substantive semantics of features, to establish what features and values there can be, and what they mean. Sometimes the first group, those working on the formal side of features, have not realized the richness of the data offered by natural languages. Conversely, those in the second group, the typologists, have not always been sufficiently concerned about the formal consequences of the patterns they have identified.

Although the theoretical interest implicit in inflectional syncretism has been recognized for some time, the range of languages that have contributed to its study has remained limited. The notion was first applied to Indo-European languages, and it is from these that the bulk of examples has been drawn. Languages from other families have been brought to bear in theoretically or typologically oriented works, but to a lesser extent, e.g. Afro-Asiatic (Carstairs 1987, Fradkin 1991, Johnston 1997, Noyer 1997), Uralic (Bátori 1990, McCreight and Chvany 1991, Kiparsky 2001, Blevins 2003), Tibeto-Burman (Gvozdanović 1991), Altaic (Carstairs 1987), and more broadly defined, languages of the Caucasus (Hjelmslev 1935–7, Boeder 1976, Carstairs 1987, Carmack 1997, Kibrik 1997, Helmbrecht 1999), of Australia (Zaliznjak 1973, Goddard 1982, Bavin and Shopen 1987, Heath 1991, Noyer 1997, Evans, Brown and Corbett 2001), of the Americas (Boeder 1976, Hewson 1989, Heath 1998, Lakämper and Wunderlich 1998, Harbour 2003), of Siberia (Kibrik 1997, Spencer 2000) and New Guinea (Kibrik 1997, Noyer 1998, Wunderlich 2001b). However, these works address at most several languages. Cross-linguistic studies relevant to syncretism are rare: Cysouw (2003) is a comprehensive treatment of the morphological and lexical expression of pronominal features (person and number), while Arkadiev (2003) discusses case syncretism.Thus, there remains a real gap in our empirical understanding of syncretism, which the following chapter should begin to redress. We have gathered examples from a wide range of languages.

The Slavonic Languages edited by Bernard Comrie and Greville G. Corbett is reviewed.

We have distinguished the groups into which nouns can be divided (controller genders) from the sets of markers (target genders) which appear on agreeing elements. Nouns are assigned normally to a single gender, while agreeing elements or targets have more than one gender form, and the selection of the appropriate form depends on the gender of the controller. It is interesting topics concerned with target tenders which are the focus of this chapter. First we take up from the preceding chapter the question of the interaction with number, and the types of syncretism which arise (section 7.1). Then we consider the form of gender agreement used when the normal conditions for agreement are not met and so gender agreement becomes a problem. This question is covered in two sections: neutral agreement in section 7.2, and gender agreement with noun phrases involving reference problems in section 7.3. Discussion of inconsistent patterns of agreement is reserved for the following chapter.Gender and numberIt has already become apparent that number enjoys a special relationship to gender. We saw examples in the last chapter where an agreement class which was based on a difference in agreement in one morphological case was not then recognized as a gender (but as a subgender); on the other hand, agreement classes based on a difference in number were recognized as genders.

This book has required a great deal of informant work and many hours of consultation with experts on particular languages and language groups. It is a pleasure to record my gratitude to all those who have been generous with their time and expertise, in providing examples or references, discussing data, or commenting on parts of the book: Jean Aitchison, Keith Allan, Gunilla Anderman, R. E. Asher, Stephen Barbour, Michael Barlow, Ruth Berman, Catherine Chvany, Ulrike Claudi, Joseph Clements, Richard Coates, N. E. Collinge, Francis Cornish, Merton Dagut, Anna Morpurgo Davies, Margaret Deuchar, R. M. W. Dixon, Donka Farkas, William Foley, Ives Goddard, Nigel Gotteri, Joseph Greenberg, Dick Hayward, Bernd Heine, Eugénie Henderson, Richard Hogg, Dee Ann Holisky, Dick Hudson, Jim Hurford, Larry Hyman, Ewa Jaworska, A. A. Kibrik, Ewan Klein, A. I. Kovaĺ, Graham Mallinson, Naomi Martin, Igor Meĺĉuk, Anne Mills, Ngessimo Mutaka, Yoni Neeman, Almerindo Ojeda, John Payne, David Perlmutter, Rebecca Posner, Malathi Rao, Bob Rothstein, Linda Schwartz, Roland Sussex, Karen Taylor-Browne, and W. A. A. Wilson. F. R. Palmer, who has read and commented on each draft chapter, deserves special thanks. Naturally, those listed do not necessarily agree with my analyses. Graphic representations are a great help in giving a clear account of some parts of the topic, and I am grateful to Ian Clark, Annie Read and Kevin Shaughnessy for artwork. Some examples with numerous diacritics almost required artwork too, so the word-processing skills of Carole D'Arcy, Pauline Rayner and Philippa Galloway were appreciated.

Mayali has four genders and five morphological classes, with formal identity between the gender prefixes and four of the morphological class prefixes. Gender and morphological class are assigned according to different but largely overlapping semantic principles. We analyze these partially overlapping systems within the NETWORK MORPHOLOGY framework; an implemented model demonstrates that the analysis gives the correct forms for the majority of nouns in a basic lexicon, and further extends to understanding assignment in the avoidance register. Our account depends on recognizing two different types of default: NORMAL CASE DEFAULT, the expected outcome in a given domain, and EXCEPTIONAL CASE DEFAULT, the last resort short of full lexical specification.

Morphosyntactic features are a difficult challenge for the typologist. It is argued that the way to distinguish canonical morphosyntactic features is through their interaction with canonical parts of speech. This is defined through four criteria. The deviations from them help to characterize several problems involved with feature systems. The main result is that number can, in certain languages, come very close to being canonical. Gender, person, and case are defined in turn as showing deviations from the fully canonical situation.

Up to this point we have assumed that nouns can be divided into genders and we have analysed the composition of these genders, considering whether they are based solely on semantic criteria, or whether formal factors also have a role. We now turn to gender agreement. This is important for two reasons: first, it is the way in which gender is realized in language use; and second, as a consequence, gender agreement provides the basis for defining gender and for establishing the number of genders in a given language. In this chapter we concentrate on the variety of ways in which gender is exemplified in the languages of the world, leaving to chapter 6 the procedures for determining the number of genders in a given language. While there is a broad consensus on the core cases of agreement, there is no generally accepted definition; there is a problem as to the outer limit of phenomena properly described as agreement, as we shall see when the personal pronoun is discussed in section 5.1. A working definition is provided by Steele (1978: 610):The term agreement commonly refers to some systematic covariance between a semantic or formal property of one element and a formal property of another. For example, adjectives may take some formal indication of the number and gender of the noun they modify.

The term ‘gender’ requires discussion, since different linguistic traditions describe gender phenomena using different terms. Moreover, while in many languages there is no dispute as to the number of genders, there are other languages where the question is far from straightforward. This requires consideration of the analytical problem of determining the number of genders in a given language. The central problem, however, is gender assignment, that is, the way in which the native speaker allots nouns to genders. Examination of languages from different families reveals that genders always have a semantic core, which may be biological sex, or animacy, with other features also having a role. In some languages nouns are assigned to genders solely on the basis of semantics, but in others this semantic information is supplemented by formal information, which may be morphological or phonological.

Since features are shared across sub-areas of the discipline, from the highly theoretical to the most applied, there are practical steps which can have general benefit. Thus there are various proposals for standardizing and generalizing the use of features and values. It is important that such initiatives should be fully informed by the linguistic issues. We shall also look at instances of using features in large-scale implementations; these serve as a valuable testing-ground for theories of features and can come to operate as de facto standards.The Leipzig Glossing RulesThe Leipzig Glossing Rules were introduced in §1.3, with source information, and they are used through the book. Although features are common currency, they can lead to confusion even at the most basic level, through the use of abbreviations in glossing which have different interpretation. The Leipzig Glossing Rules represent a bottom-up approach to standardization and are a useful step forward. They give conventions for glossing, and propose some standard abbreviations. At the simplest level it is eminently sensible that we should use the same symbols (e.g. ‘=’ for clitic boundary) and the same abbreviations. We should be certain whether a colleague wishes to indicate perfect tense (prf) or perfective aspect (pfv). That is, we should disagree on real issues, not be sidetracked by gratuitous differences in labelling. This is a bare minimum for the discipline. Interestingly, at the annual meeting of the Societas Linguistica Europaea (2011) it was specified that the Leipzig Glossing Rules should be used in handouts, which is one sign of progress being made.

As we saw in the previous chapter, the challenge for a formal analysis of syncretism is twofold: to represent the cross-linguistically more general patterns of syncretism in terms of feature structure, and to account for patterns which occur independently of feature structure. In this chapter we lay out a formal framework and demonstrate it with three cases studies. When introducing the formal framework, we show in §5.1 that inferential-realizational frameworks, such as Network Morphology and Paradigm Function Morphology, are capable of capturing syncretic patterns which may arise as the result of underspecification, or as the result of referrals (i.e. independently of feature structure). One advantage of such approaches is that referrals and underspecification can be used simultaneously. We shall see when we come to the second case study that this is just what is required for the avoidance morphology of Dalabon. Indeed, we show that generalized referrals – where sets of paradigm cells can refer to sets of paradigm cells – which frameworks such as Network Morphology allow for, are an important requirement.We consider the relationship between underspecification (a feature structure-based constraint) and semantic naturalness. As we saw in Chapter 4: §4.2.1, underspecification can be used with atomic feature values (i.e. in a ‘flat’ structure) where the syncretism is the default to the ‘elsewhere’. In such cases, in the absence of other representational means, only one syncretic pattern can be described for any domain.

Heads in grammatical theoryThe majority of current grammatical theories refer explicitly to the head of a phrasal constituent. Yet while the term ‘head’ has entered the common currency of theoretical linguistics, this does not provide evidence of agreement on what it means. Nor does the term's long and varied career in linguistics guarantee that it identifies a notion which is not already identified by some other, more basic notion or interacting set of notions. The purpose of this volumev is twofold: first, it aims to uncover and make explicit the notion (or notions) behind the term ‘head’; second, it aims to investigate the status of the notion (or notions) in linguistic theory.Most linguists would agree with the informal characterization that the head of a phrase is one of its constituents which in some sense dominates and represents the whole phrase. In an important paper published in the Journal of Linguistics in 1985, Zwicky drew attention to the fact that use of the term ‘head’ had been extended from syntax to morphology (for example, by Lieber, 1981; Williams, 1981; and Kiparsky, 1982) in spite of the fact that there was no generally agreed formal definition of the notion in syntax (though an important contribution had already been made by Gazdar and Pullum, 1981 and Gazdar, Pullum and Sag, 1982). Zwicky therefore set out to find a rigorous, generally acceptable definition for ‘head’. He proceeded by examining the following eight candidate criteria for the identification of a constituent as a syntactic head.

This book has an interesting history of collaboration. It began life in research done by Greville Corbett and Norman Fraser on the morphology of Russian, starting in 1990, research which was inspired by the work of Roger Evans and Gerald Gazdar on DATR. The ESRC and Leverhulme Trust provided funding, which brought Dunstan Brown and Andrew Hippisley to Surrey, and the work developed into a more general theoretical framework, Network Morphology. We found syncretism of increasing importance in the development of the framework and gave presentations at the following places: Krems (Austria), University of Sussex, Linguistics Association of Great Britain (at the University of Surrey), University of California (Berkeley), Gregynog (Wales), Heinrich-Heine-Universität (Düsseldorf), University of Edinburgh, University of Cologne, University of Helsinki, La Trobe University, Norsk Forening for Språkvitenskap (Oslo), Institutt for Østeuropeiske og Orientalske Studier (University of Oslo), Moscow University, University of Oxford, Cornell University, Twelfth International Conference on Historical Linguistics (University of Manchester), Conference on Lexical Structures (Wuppertal), British Association for Slavonic and East European Studies (Cambridge), University of Sheffield, University of Essex, University of Pennsylvania, Leipzig University, Association for Linguistic Typology (University of Amsterdam), Second Mediterranean Meeting on Morphology (University of Malta), Max Planck Institute for Psycholinguistics (Nijmegen), Second Winter Typological School (Istra, Moscow district), Ninth International Morphology Meeting (Vienna), University of California (Santa Barbara), University College London, Second Northwest Conference on Slavic Linguistics (Berkeley), Stockholm University, 37th Meeting of the Chicago Linguistic Society, Fakultetets Forsknings Fredage (University of Copenhagen), University of Melbourne, Scandinavian Slavists' Summer School (Kungälv, Sweden), University of Leeds, School of Oriental and African Studies (London), University of Manchester Institute of Science and Technology, Lancaster University, University of Catania, Second International Seminar ‘Computer Treatment of Slavonic Languages’ (Bratislava), and University of York.

In chapters 2 and 3 we examined data from a wide range of languages and established a typology of gender systems, based on the type of criteria by which nouns are allotted to genders. We also saw that there can be considerable overlaps; a noun may be, say, feminine because of its meaning, morphology and phonology. The question then arises as to which factors are actually used by native speakers. We must therefore ask what is the evidence for the psychological reality of the gender assignment systems discussed. The major evidence is, of course, the data already presented. Given the massive regularities established, and the ease with which native speakers use gender, the most plausible explanation is that speakers assign nouns to genders without difficulty simply by taking advantage of these regularities. We now turn to other facts which help confirm that assignment rules are indeed part of the native speaker's competence and not just regularities observed by linguists. The first type of evidence is provided by borrowings (section 4.1); as new nouns are borrowed into a language they must be given a gender and this allows us to see the assignment rules operating on material which is sometimes unlike that of the native vocabulary. Then we should consider how children acquire gender (section 4.2), as they may learn parts of the system before others and so confirm that there are separate factors at work.

We report a study of the incidence of ‘colour-blindness’ in southern and central Africa, and we compare the African data with data from various European groups. There was a surprisingly high incidence of tritan errors (yellow–blue defect). The likelihood of making tritan errors increased with age, and was greater in rural areas than in towns. In Europe, no tritan errors were made by samples from the U.K., Eire or Spain, but some tritan errors were made by a sample from southern Greece. In contrast, most of a British sample of people over sixty-five years old makes tritan errors. Although tritan errors were the most frequent, they were often accompanied by protan and deutan errors. This mixed pattern of errors is consistent with the condition being acquired rather than congenital. Many languages of southern Africa categorise blues and greens with the same term. If the tritanopia we report has been endemic, it may have reduced the ‘perceptual pressure’ to split the blue-with-green categories into separate blue and green terms; a speculation consistent with Rivers, W. H. R. (1901. Introduction to A. C. Haddon (Ed.), Reports on the Cambridge Anthropological Expedition to the Torres Straits. Cambridge: Cambridge University Press) © 1998 Elsevier Science Ltd. All rights reserved..

In this chapter we tackle a most interesting problem, namely that of hybrid nouns. As early as Chapter 3 we noted cases where the meaning and the form of nouns conflict in terms of gender assignment. Normally semantic criteria overrule formal considerations. In some instances, however, the conflict of criteria is not settled in this unambiguous way, and a hybrid noun results. The specific nature of hybrid nouns was identified in section 6.4.5.2. Like nouns of double (or multiple) gender, hybrid nouns take more than one set of agreements, that is, they take forms from more than one consistent agreement pattern. But unlike nouns of double gender, hybrid nouns do not simply belong to two genders. The crucial point about hybrid nouns is that the form of gender agreement used with them depends in part on the type of agreement target involved. Thus, while we can say of a normal noun simply that it takes, for example, feminine agreement, for a hybrid noun we can specify the agreement only provided we know the agreement target in question. Given this, the range of possible inventories of agreements taken by hybrids would appear to be extensive. However, we shall see that there are generalizations to be made about such agreement options; they are constrained by the Agreement Hierarchy (section 8.1). It turns out that pronouns have a special importance for hybrid nouns, which is examined in section 8.2.

In this chapter we concentrate on the possible meaning distinctions in number systems. Often the situation in languages like English is taken as normal, whereas it represents only one of the possibilities. We will first consider whether number needs to be expressed; we shall see that for some languages the expression of number is in a sense optional, while in others it is a category which speakers cannot avoid. To investigate these systems we shall first consider the notion of ‘general’ number as a meaning distinction and base a partial typology upon it (§2.1). We then narrow our attention to the cases where number is expressed, and establish the main types of distinction within the category (§2.2). Thus §2.1 is devoted to the opposition of number and ‘non-number’, while §2.2 examines the possibilities within the number domain. In §2.3 we propose a typology, systematizing the material examined so far, and we go on to show that languages may simply not have a number system (§2.4); then we consider approaches to number within formal semantics (§2.5).Our aim in this chapter is to find all the possible distinctions. At this stage we shall not be concerned about the type of nominal we look at, so long as we find those which show the greatest differentiation. Keeping any particular nominal ‘still’ as it were, we shall see how many different numbers it may have available, in the most favourable contexts.

"Nominal classification in Aboriginal Australia" edited by Mark Harvey and Nicholas Reid is reviewed.

Research into entries and rules based on English is hampered by the fact that crucial factors coincide. Clahsen avoids this problem and demonstrates his claims by working on German. Seen against the background of the immense variety of the world's languages, this successful move is potentially the first of many possible ones: Several languages offer promising configurations of the factors relevant to Clahsen, and others present new challenges.

The canonical approach is designed to avoid two dangers in typology: the premature use of statistics and the failure to compare like with like. It involves taking definitions to their logical end point, in order to build theoretical spaces of possibilities. The canonical instances are not necessarily the most frequent, and may indeed be rare, but they fix a point from which occurring phenomena can be calibrated. This approach is demonstrated first in syntax, by looking at agreement. It is then applied to morphology, starting from the notion of a canonical paradigm, and showing how four major morphological phenomena can be seen as mismatches between the canonical expectation and the variety of real inflectional systems. Syncretism and suppletion are investigated as examples. The latter demonstrates how typology can be applied to an extreme case (lexical items whose constituent parts have no shared phonology), which can be extended by interactions with other morphological phenomena. The relevance of this approach is shown in its application to the construction of typological databases.

Gender is the most puzzling of the grammatical categories. It is a topic which interests non-linguists as well as linguists and it becomes more fascinating the more it is investigated. In some languages gender is central and pervasive, while in others it is totally absent. One of its attractions for linguists is that there are interesting aspects of the study of gender in each of the core areas of linguistics. And work on it promises practical benefits, even in the short term, in meeting the problems which gender causes in second-language learning. In the longer term, research into gender will be important for at least two other areas: first, it can shed light on the way in which linguistic information is stored in the brain; and second, it has implications for natural language processing, notably for the elimination of local ambiguities in parsing. To understand what linguists mean by ‘gender’, a good starting point is Hockett's definition: ‘Genders are classes of nouns reflected in the behavior of associated words’ (1958: 231). A language may have two or more such classes or genders. The classification frequently corresponds to a real-world distinction of sex, at least in part, but often too it does not (‘gender’ derives etymologically from Latin genus, via Old French gendre, and originally meant ‘kind’ or ‘sort’). The word ‘gender’ is used not just for a group of nouns but also for the whole category; thus we may say that a particular language has, say, three genders, masculine, feminine and neuter, and that the language has the category of gender.

In the last chapter we examined languages in which gender is assigned solely by semantic criteria. We also noted languages in which semantic criteria allowed various numbers of exceptions. We now come to languages in which large numbers of nouns fall outside the semantic assignment rules. These nouns may be handled instead by formal assignment rules, that is, rules which depend on the form of the nouns involved rather than on their meaning. These rules are of two types, morphological and phonological, which we will consider in turn (sections 3.1 and 3.2). Whereas the distinction between semantic and formal assignment rules is clear (though their effects may overlap), the distinction between morphological and phonological rules is not always clear-cut. As a rule of thumb, we may say that phonological rules refer just to a single form of a noun, for example, ‘nouns ending in a vowel in the singular are feminine’. Typically, the most basic form of the noun is involved, though this is not always spelled out. Morphological rules, on the other hand, require more information; they need to refer to more than one form. This is not always obvious. A typical assignment rule of the morphological type might be: ‘nouns of declension II are feminine’; establishing that a noun is of declension II might require information about, say, the nominative singular and the genitive singular. Note that there are no syntactic systems; the obvious syntactic system would be one in which gender was assigned to nouns according to agreement.

The lexicon divides into parts of speech (or lexical categories) and there are cross-cutting regularities (features). These two dimensions of analysis take us a long way, but several phenomena elude us. For these the term ‘split’ is used extensively (‘case split’, ‘split agreement’, and more), but in confusingly different ways. Yet there is a unifying notion here. I show that a split is an additional partition, whether in the part of speech inventory or the feature system. On this base an elegant typology can be constructed, using minimal machinery. The typology starts from four external relations (government, agreement, selection, and anti‑government), and it specifies four types of split within each (16 possibilities in all). This typology (i) highlights less familiar splits, from diverse languages, and fits them into the larger picture; (ii) introduces a new relation, anti‑government, and documents it; (iii) elucidates the complexities of multiple splits; (iv) clarifies what exactly is split, which leads to a sharpening of our analyses and applies across different traditions.

This book presents a controlled evaluation of three widely practised syntactic theories on the basis of the extremely complex agreement system of Archi, an endangered Nakh-Daghestanian language. Even straightforward agreement examples are puzzling for syntacticians because agreement involves both redundancy and arbitrariness. Agreement is a significant source of syntactic complexity, exacerbated by the great diversity of its morphological expression. Imagine how the discipline of linguistics would be if expert practitioners of different theories met in a collaborative setting to tackle such challenging agreement data - to test the limits of their models and examine how the predictions of their theories differ given the same linguistic facts. Following an overview of the essentials of Archi grammar and an introduction to the remarkable agreement phenomena found in this language, three distinct accounts of the Archi data examine the tractability and predictive power of major syntactic theories: Head-driven Phrase Structure Grammar, Lexical Functional Grammar, and Minimalism. The final chapter compares the problems encountered and the solutions proposed in the different syntactic analyses and outlines the implications of the challenges that the Archi agreement system poses for linguistic theory.

Agreement in language occurs when grammatical information appears on a word which is not the source of that information. In simple examples like she runs, the form runs is singular, agreeing in number with she. This is information about the number of runners (just one), and it matches that expressed in its source she. Patterns of agreement vary dramatically cross-linguistically, with great diversity in expression and types of variation found. This clear introduction offers an insight into how agreement works, and how linguists have tried to account for it. Comparing examples from a range of languages, with radically different agreement systems, it demonstrates agreement at work in a variety of constructions. It shows how agreement is influenced by the conflicting effects of sentence structure and meaning, and highlights the oddities of agreement in English. Agreement will be essential reading for all those studying the structure and mechanisms of natural languages.

CLS37: The Panels. Papers from the 37th Meeting of the Chicago Linguistic Society, 15-28

The databases record instances of deponency, which is the term we have adopted to describe mismatches between morphology and morphosyntax. The prototypical example are the deponent verbs of Latin, which involve a mismatch between passive form and active meaning. That is, a normal Latin verb had active forms such as amō 'I love' and amāvī 'I have loved', which contrasted with the passive forms amor 'I am loved' and amātus sum 'I have been loved' (in this case, with a masculine subject). A deponent verb, on the other hand, looks like the passive but functions like the active, as in mīror 'I admire', mīrātus sum 'I have admired'. In the the databases we construe deponency in an extended fashion, covering any mismatch between the apparent morphosyntactic value of a morphological form and its actual value in a given syntactic context. Two databases are housed on this site, accessible through the links above. The cross-linguistic database looks at the presence of morphological mismatches in a controlled sample of genetically and geographically diverse languages (based on the 100-language sample from the World Atlas of Language Structures). The typological database records the logical space of deponency: what features may be affected, and what are the characteristics of the resulting paradigm? Every logical combination of parameters is represented by one exemplar (or where none has been found, this is noted too). The typological database is supplemented by a set of formal analyses of examples which hold particular interest for morphological theory.

paper presented at SLOVKO 2003 second international seminar on computer treatment of Slavonic languages. Bratislava, Slovakia 24-5 October 2003

This revised edition includes updated bibliographies for each chapter and up-to-date census figures. The featured languages have been chosen based on the number of speakers, their role as official languages and their cultural and historical importance. Each language is looked at in depth, and the chapters provide information on both grammatical features and on salient features of the language's history and cultural role.

In this chapter we consider databases which have been constructed to investigate particular linguistic phenomena. Data entered into a database with little thought or attention to its categorisation are at best usless, and in the worst case harmful if used to make spurious generalizations. So there is a requirement that we are explicit about our analyses and the phenomena under investigation, and that serious thought is given to the structure of the database.

We present a new phenomenon in inflectional morphology, ‘repartitioning’, based on data from Soq (Trans New Guinea). In repartitioning, the semantic boundary between two sets of morphological forms is redrawn in a single domain; one feature value takes over part, but not all, of the meaning of the other. In Soq the boundary is redrawn between the yesterday past tense and the hodiernal; the domain is the lexeme s- ‘stay’. For this one verb, the yesterday past takes over most of the range of the hodiernal, while the morphological forms remain regular. In clause chains the repartitioned verb surprisingly shows no syntactic effects. We demonstrate key differences from known phenomena, notably syncretism and overdifferentiation. Repartitioning is indeed new. It can be modeled in a theory based on default inheritance, but poses problems for other approaches. Finally, we present a typology of featural mismatches that situates Soq relative to known phenomena.*

Many languages have an intriguing class of nouns, the pluralia tantum, which have restricted number possibilities when, in some sense, they should not. Thus English binoculars has no singular, which is worth noting (that is, it is not predictable). True, there are other nouns denoting items consisting of two significant parts which behave similarly (spectacles, trousers ...); indeed they are subject to ‘middle-size generalizations’ (Koenig 1999). But there are two reasons to note such nouns. First there are many English nouns equally denoting items consisting of two significant parts which are unremarkable in this respect: bicycle, bigraph, Bactrian camel, couple, duo ... And second, there are languages with number systems roughly comparable to that of English in which the equivalents of binoculars and trousers are normal count nouns: Russian binokl´, French pantalon. While pluralia tantum are of continuing interest, it is typically only the English type which is considered. But these familiar examples offer an entry point to a collection of lexical items, some with much stranger behaviour, lurking between the semi-predictable and the unexpectedly defective. In particular, some instances demonstrate that we cannot maintain the general assumption that the ‘internal’ morphosemantic specification of a paradigm cell and its ‘external’ morphosyntactic requirement are necessarily identical. I therefore set out a full typology of these fascinating nouns, so that their significance can be more fully appreciated. I start from the notion of canonical noun, and calibrate the different noncanonical properties according to a set of orthogonal criteria.

As Einstein nicely put it: ‘Everything should be made as simple as possible, but not simpler.’ It would be good to have a simple typology of the morphosyntactic features. And if Zwicky (1986: 988-989) is right in suggesting that there is a fixed list of available features and values, then a simple typology is an attainable goal. However, when we examine how features and their values can be established for given languages, gradience appears to be a major challenge. One common response to this challenge is to propose additional feature values; as we shall see, this response would rule out a simple typology. I argue that this approach fails: additional values cannot account for gradience. This means that we can still work towards the ideal of a simple typology. Thus for this paper, gradience looms large as a possible obstacle to a different goal. Gradience is an obstacle, which might suggest that we were heading for the ‘simpler’ situation, the one that fails to recognize the true complexity of the problem. I will show that this apparent obstacle is not directly relevant, allowing us still to work towards a typology which is indeed ‘as simple as possible’

Gender distinction is often based on sex; sometimes this is only one criterion and the gender of nouns depends on other factors (thus "house" is masculine in Russian, feminine in French and neuter in Tamil).

The Papuan language Mian allows us to refine the typology of nominal classification. Mian has two candidate classification systems, differing completely in their formal realization but overlapping considerably in their semantics. To determine whether to analyse Mian as a single system or concurrent systems we adopt a canonical approach. Our criteria – orthogonality of the systems (we give a precise measure), semantic compositionality, morphosyntactic alignment, distribution across parts of speech, exponence and interaction with other features – point mainly to an analysis as concurrent systems. We thus improve our analysis of Mian and make progress with the typology of nominal classification.

The abstract structure of paradigms reflects the requirements of syntax. This implies regularity across the lexicon. How then can the structure of a paradigm (rather than particular forms) be lexicalized? There are possible instances of this in Slavonic, which we discuss both for their inherent interest and for their relevance to the more general issue of the syntax-morphology interface. Our approach relies heavily on the notion of default, and is within the Network Morphology framework (Corbett & Fraser 1993, Evans, Brown & Corbett 2002, Brown & Hippisley 2012). In this paper the emphasis is on the interesting conceptual issues, rather than on the formalism.

One of the milestones in typological studies is Berlin & Kay's (1969) account of basic colour terms, which has produced a steady stream of research of various types. Berlin & Kay summarized their work as follows. In sum, our two major findings indicate that the referents for the basic color terms of all languages appear to be drawn from a set of eleven universal perceptual categories, and these categories become encoded in the history of a given language in a partially fixed order (1969: 4–5).

Number is the most underestimated of the grammatical categories. It is deceptively simple yet the number system which philosophers, logicians and many linguists take as the norm - namely the distinction between singular and plural (as in cat versus cats) - is only one of a wide range of possibilities to be found in languages around the world. Some languages, for instance, make more distinctions than English, having three, four or even five different values. Adopting a wide-ranging perspective, Greville Corbett draws on examples from many languages to analyse the possible systems of number. He reveals that the means for signalling number are remarkably varied and are put to a surprising range of special additional uses. By surveying some of the riches of the world’s linguistic resources this book makes a major contribution to the typology of categories and demonstrates that languages are much more varied than is generally recognised.

The author uses a 'canonical' approach to offer a new perspective on the complex phenomenon of inflectional classes. This means extrapolating from what there is to what there might be, in order to define the theoretical space into which real instances fit. To do this, the author proposes eight criteria, grouped under two overarching principles. These are: I. distinctiveness: canonical inflectional classes are as clearly distinct as possible; and II. independence: the distribution of lexical items over canonical inflectional classes is unmotivated. The author investigates the various deviations from these principles, by considering in turn the more detailed criteria which exemplify them. While one might reasonably expect that 'canonical inflectional class' is an ideal without exemplars, the author finds an example which comes remarkably close to canonical.

Cross-linguistic database and typological database

There are fascinating problems at the syntax-morphology interface which tend to be missed. I offer a brief explanation of why that may be happening, then give a Canonical Typology perspective, which brings these problems to the fore. I give examples showing that the phenomena could in principle be treated either by syntactic rules (but these would be complex) or within morphology (but this would involve redundancy). Thus `non-autonomous' case values, those which have no unique form but are realized by patterns of syncretism, could be handled by a rule of syntax (one with access to other features, such as number) or by morphology (with resulting systematic syncretisms). I concentrate on one of the most striking sets of data, the issue of prepositional government in Latvian, and outline a solution within Network Morphology using structured case values.

We examine the varying role of conditions on grammatical relations marking (namely animacy and volitionality) by looking at different languages of one family, using both existing descriptions and working with specially prepared video stimuli. This enables us to see the degree of variation permitted within closely related languages. We look at four Alor-Pantar languages (Teiwa, Adang, Kamang, and Abui), Papuan languages of eastern Indonesia. The conditions on argument marking are manifested in different ways. Those languages with syntactic alignment index objects with a prefix, those which have semantic alignment index objects and some subjects with a prefix. In 42 video clips we systematically varied animacy and volitionality values for participants in one and two-participant events. These clips were used in fieldwork to elicit descriptions of the events. The data show that animacy of the object is an important factor which favours indexation of the object on the verb in all four languages to varying degrees. Volitionality, on the other hand, is a factor in the semantically aligned languages only. While the presence of a prefix on the verb is semantically motivated in many instances, marking is not directly determined by verbal or participant semantics, and lexical factors must also play a role.

The alveolar fricative occurs in word-final position in English in different grammatical functions. Nominal suffixes may indicate plurality (e.g. cars), genitive case (e.g. car’s) or plurality and genitive case in cumulation (e.g. cars’). Further, there are the third person singular verbal suffix (e.g. she fears) and the cliticized forms of the third person singular forms of have and be (e.g. she’s been lucky; she’s friendly). There is also non-affixal s (e.g. freeze (noun)). Against the standard view that all these types are homophonous, several empirical studies have shown that at least some of the fricatives listed can actually be differentiated in their duration. The present article expands this line of research and considers a further case, which has not been included in previous analyses: pluralia-tantum nouns (e.g. goggles). We report on a carefully controlled reading study in which native speakers of British English produced pluralia-tantum and comparable regular-plural nouns (e.g. toggles). The duration of the word-final fricative was measured, and it was found that the two do not systematically differ in this acoustic parameter. The new data are interpreted in comparison to relevant previous studies, and against the background of the similarities of pluralia-tantum and regular-plural nouns.

Competition takes many forms. A newly identified type of competition involves the featural specification of one of the competitors as a key factor. In the particular instance treated here, whether a given item has a competitor depends on its number (and sometimes its person). We focus on the use of the genitive case versus adjective-like forms in possessive expressions (broadly understood). The data come primarily from the Slavonic languages, where a surprising original system of possessive pronouns competing with personal pronouns has played out rather differently through the family. We find a variety of outcomes, from conservative to highly innovative, with some instances of competitors settling into different niches.

The term 'gender' requires discussion, since linguistic traditions differ here. This requires us to confront the analytical issue of determining the number of genders in a given language. The central concern of this article will be gender assignment-that is, the way in which the native speaker allots nouns to genders. Examination of languages from different families reveals that genders always have a semantic core, which may be biological sex, or animacy, with other features also having a role. In some languages, nouns are assigned to genders solely on the basis of semantics, but in others this semantic information is supplemented by formal information, which may be phonological or morphological. Given this typology, it is apparent that gender is distributed in interesting ways across the world's languages. Finally, prospects for research into gender are considered. © 2006 Elsevier Ltd. All rights reserved.

Syncretism is a surprising yet widespread and poorly understood phenomenon in natural language. Given a regular distinction such as present versus past, as in English help/helped, work/worked, laugh/laughed, we might not expect to find instances like bid, which can be present or past (we now bid five pounds, though yesterday we bid ten pounds). The form bid, is said to be an instance of syncretism, a single form fulfilling two or more different functions. Thus syncretism is found even in English, whose inflectional morphology (system of different word-forms) is simple in comparison with many languages. The database encodes information on inflectional syncretism in 30 genetically and geographically diverse languages, representing such morphosyntactic features as case, person, number and gender, in all the inflectional classes where they are relevant.

Offers new perspectives on basic elements of linguistic analysis Subject is of interest to all linguists Important for both theoretical and empirical research Written by prominent, international scholars This book presents a critical overview of current work on linguistic features and establishes new bases for their use in the study and understanding of language. Features are fundamental components of linguistic description: they include gender (feminine, masculine, neuter); number (singular, plural, dual); person (1st, 2nd, 3rd); tense (present, past, future); and case (nominative, accusative, genitive, ergative). Despite their ubiquity and centrality in linguistic description, much remains to be discovered about them: there is, for example, no readily available inventory showing which features are found in which of the world's languages; there is no consensus about how they operate across different components of language; and there is no certainty about how they interact. This book seeks at once to highlight and to tackle these problems. It brings together perspectives from phonology to formal syntax and semantics, expounding the use of linguistic features in typology, computer applications, and logic. Linguists representing different standpoints spell out clearly the assumptions they bring to different kinds of feature and describe how they use them. Their contrasting contributions highlight the areas of difference and the common ground between their perspectives. The book brings together original work by leading international scholars. It will appeal to linguists of all theoretical persuasions. Readership: Linguists of all theoretical persuasions - including syntacticians, morphologists, computational linguists, and typologists - and their postgraduate students.

We describe a quick and robust procedure for establishing likely basic colour terms. We illustrate the procedure with a study of English where the basic colour term inventory is known. The method consists of two tasks: an elicited list task (tell me as many colour terms as you know) and a colour naming task. The list task elicits contenders for the basic colour term slots and the naming task establishes their range of referents. The indicators of the degree of basicness of colour terms converged to confirm that there are eleven basic colour terms in English.

It is a pleasure to acknowledge my intellectual debt to Adam Evgen'evič; I have learned a good deal from his careful and extensive work on numerals in different Slavonic languages. A natural question arising from that work is the way in which numerals and grammatical number interact, a question we shall consider in section 2. Adam Evgen'evič's research shows that the interaction is somewhat surprising (section 3), specifically that noun phrases with higher numerals are less likely to control plural agreement than are those with lower numerals. We consider the implications of this finding in terms of individuation (section 4). And then we go on to show how in the light of the Slavonic situation we can better understand puzzling data from languages unrelated to Slavonic, namely Bayso and Arabic (section 5).

Paper in the Arbeitsgruppe 'Auf alles gefasst sein: Ausnahmen in der Grammatik' at the 27th Annula meeting of the Deutsche Gesellschaft fur Sprachwissenschaft, Cologne 23-25 Feb 2005

An important design feature of language is the use of productive patterns in inflection. In English, we have pairs such as 'enjoy' ~ 'enjoyed', 'agree' ~ 'agreed', and many others. On the basis of this productive pattern, if we meet a new verb 'transduce' we know that there will be the form 'transduced'. Even if the pattern is not fully regular, there will be a form available, as in 'understand' ~ 'understood'. Surprisingly, this principle is sometimes violated, a phenomenon known as defectiveness, which means there is a gap in a word's set of forms: for example, given the verb 'forego', many if not most people are unwilling to produce a past tense. Although such gaps have been known to us since the days of Classical grammarians, they remain poorly understood. Defectiveness contradicts basic assumptions about the way inflectional rules operate, because it seems to require that speakers know that for certain words, not only should one not employ the expected rule, one should not employ any rule at all. This is a serious problem, since it is probably safe to say that all reigning models of grammar were designed as if defectiveness did not exist, and would lose a considerable amount of their elegance if it were properly factored in. This volume addressed these issues from a number of analytical approaches - historical, statistical and theoretical - and by using studies from a range of languages.

Deponency is a mismatch between form and function in language that was first described for Latin, where there is a group of verbs (the deponents) which are morphologically passive but syntactically active. This is evidence of a larger problem involving the interface between syntax and morphology: inflectional morphology is supposed to specify syntactic function, but sometimes it sends out the wrong signal. Although the problem is as old as the Western linguistic tradition, no generally accepted account of it has yet been given, and it is safe to say that all current theories of language have been constructed as if deponency did not exist. In recent years, however, linguists have begun to confront its theoretical implications, albeit largely in isolation from each other. There is as yet no definitive statement of the problem, nor any generally accepted definition of its nature and scope. This volume brings together the findings of leading scholars working in the area of morphological mismatches, and represents the first book-length typological and theoretical treatment of the topic. It will establish the important role that research on deponency has to play in contemporary linguistics, and set the standard for future work.

Languages change by gaining and losing word forms over time, but an equally significant role in their history is played by subtle shifts in the function of existing forms. Investigating such developments requires us to analyse patterns of use in large amounts of historical data, but such data are simply unavailable for most languages. Russian is a happy exception. It is a language with a rich and relatively stable system of inflectional morphology. Yet while the system of forms has changed relatively little, the use of these forms has undergone a remarkable degree of change over the last 200 years, a period for which a substantial quantity of varied material is available. The database is the product of a project funded by the Arts and Humanties Research Council (grant number RG/AN4375/APN18306); a full list of project outputs may be found at http://www.surrey.ac.uk/LIS/SMG/STMC/ By investigating a corpus of literary texts created between 1801 and 2000 (10 million words in total), we have shown how dramatically a language can change even as the actual word forms remain unchanged. The database was designed to help address two theoretical questions: • What is the nature of morphosyntactic change in a language whose morphological system remains stable? • What factors condition the choice between competing forms? The database provides statistical analyses of the competition between grammatical forms for six morphosyntactic phenomena within equal time periods, described below. We give the user the means to investigate morphological, syntactic, stylistic and socio-linguistic factors involved in historical change , and so to observe how innovative usage spreads across contexts. Besides the results of this original study, we also give the results of earlier, less complete, studies by other scholars. An annotated bibliography of these sources is available at http://www.surrey.ac.uk/LIS/SMG/STMC/Bibliography.htm The data are the result of several person-years of effort; we have published some of the findings, and we welcome further use of the database by other researchers. We want the database to be accessible to historical linguists with no knowledge of Russian, as well as to Russianists, and so we give the examples in transliterated form. The database available at http://www.smg.surrey.ac.uk/STMC/(S(qhyfve45odjjhv45tllm0sbd))/index.aspx

Linguistics, and typology in particular, can have a bright future. We justify this optimism by discussing comparability from two angles. First, we take the opportunity presented by this special issue of Linguistic Typology to pause for a moment and make explicit some of the logical underpinnings of typological sciences, linguistics included, which we believe are worth reminding ourselves of. Second, we give a brief illustration of comparison, and particularly measurement, within modern typology.

In many respects the agreement systems of Slavonic languages are close to canonical. Controllers of agreement are often present, they have overt expression of features, and they take consistent agreements. The target has obligatory bound expression of agreement, and there is matching of features values (for person, number and gender). However, Slavonic also shows several very interesting instances of agreement choices, induced by a range of different controller types. These agreement choices provide good evidence for the constraints of the Agreement Hierarchy and the Predicate Hierarchy, as well as for various types of condition on agreement, notably animacy and precedence.

A typology of gender resolution is established, followed by a typology of gender assignment. It is then demonstrated that there is an implicational link between the two: the type of resolution system found in a given language is predictable in part from the assignment system.

Eva Hajicova (ed) Abstracts: XVII International Congress of Linguists: Prague, Czech Republic, July 24-9, 2003

Periphrasis straddles the boundary between syntax and morphology, and so creates analytical and theoretical problems. These concern the nature of the word, the interaction between syntax and morphology, and the possible sizes and shapes of inflectional paradigms. Progress has been hampered because the theoretical devices available have been inadequate and because the range of data considered has been narrow. We are therefore undertaking a typological survey, adopting a ‘canonical’ approach in which we specify the different dimensions along which concrete instances of periphrasis can be classified as more or less canonical. Periphrastic constructions are common, yet relatively little is known about their typological range. Following Haspelmath (2000), our canonical approach encompasses not just verbs but also comparable phenomena involving other parts of speech (‘analytic’ forms in some traditions). Thus we include less familiar examples, such as this one, showing periphrastic expression of number on nouns. We give just two of seven cases: Nenets ti ‘reindeer’ (Ackerman 2000:3) SINGULAR DUAL PLURAL NOMINATIVE ti tex°h tiq DATIVE ten°h tex°h nyah tex°q Number and case are expressed synthetically for most cells of the paradigm, except for certain cells in the dual (represented here by the dative). We plan a careful examination of four key languages, Archi, Bininj Gun-Wok, Nenets and Sanskrit, each of which has significant and different instances of periphrasis. Based on this survey, we will establish: • the set of criteria for periphrasis; this is to define the space across which to calibrate any attested periphrastic construction. We shall start from the definitions in Ackerman & Stump (2004: 125-147). • the typological range of periphrasis, taking into account both the constructions and how they interact with language-specific morphological and syntactic rules • the diachronic stages that the syntactic phrase goes through on its way to becoming part of the morphological paradigm, and the conditions that permit this. Periphrasis can be understood in terms of a configuration of properties of canonical inflectional morphology (e.g. obligatoriness) and canonical syntax (e.g. independence of expression), as in our Nenets example. The canonical approach with its inclusive orientation allows us to calibrate and clarify the differing intuitions of typologists, and place periphrasis within the larger space of morphology-syntax interaction. Ackerman, Farrell. 2000. Lexical constructions: Paradigms and periphrastic expressions. Paper read at the LFG Workshop on Morphosyntax, Berkeley. Ackerman, Farrell & Gregory Stump. 2004. Paradigms and periphrastic expression: A study in realization-based lexicalism. Projecting Morphology, ed. by Andrew Spencer & Louisa Sadler, 111-157. Stanford: CSLI. Haspelmath, Martin. 2000. Periphrasis. Morphology: An international handbook on inflection and word formation, ed. by G. Booij, C. Lehmann & J. Mugdan, 654-664. Berlin: de Gruyter.

We present a corpus-based study of variation in case assignment of the direct object of negated verbs in Russian over the past 200 years. Superficially the system of case forms available over this relatively short period has remained largely the same, but the way in which certain cases are used has been radically altered. This is particularly apparent in the treatment of the direct object of negated verbs. We argue that various semantic factors have been involved in bringing about this change, and that the role and significance of these factors has been changing over the period under investigation. This has implications for our understanding of the role of semantics in case assignment.

Lexemes may be split internally, by phenomena such as suppletion, periphrasis, heteroclisis and deponency. Generalizing over these phenomena, which split a lexeme’s paradigm, we can establish a typology of the possible internal splits. There are also lexemes whose external requirements are split: they induce different agreement, for instance. Again, a typology of these splits has been proposed. The next logical step is to attempt a typology of the possible relations between internal and external splits. This is not straightforward, since we need to avoid spurious linkages. Four lines of argument are offered: (i) general plausibility: the internal-external linkage is compelling, and so other accounts require a degree of coincidence which is unlikely; (ii) overabundance: alternative inflectional forms link to different external requirements; (iii) variation in time and space: splits in inflection and in external requirements vary, while maintaining their linkage; (iv) pluralia tantum nouns: the different types of these nouns provide intriguing confirming evidence. Case studies include Asia Minor Greek, Polish, Russian, Scottish Gaelic, Sɛlɛɛ, Serbo-Croat (BCMS), Slovenian, Latin and Old Frisian. The clear instances which emerge, where an external split is demonstrably linked to an internal one, prove both surprising and significant. We discover that in split paradigms, besides overt overabundance, there may also be covertly overabundant cells. Furthermore, when external splits involve individual cells, these will not induce simple (consistent) agreement. This makes good sense, demonstrating that featural information is associated with lexemes in a natural default manner: at the lexeme level by default, unless overridden at the sub-paradigm level, unless in turn overridden at the level of individual cells. Some lexical items show remarkable properties: they may lack internal consistency, they may be externally inconsistent and – in a fascinating minority – these two characteristics may be linked. This last type is rare, but it deserves special attention for what it tells us about the nature of lexical information and the way this is stored. I first review briefly the typology of lexeme-internal splits; these reflect phenomena such as suppletion, periphrasis, heteroclisis and deponency, and they reveal higher-level patterns. Then I summarize the types of external splits (notably different case and agreement requirements), and the conditions on such splits. This outline of internal and external splits forms the essential basis for analysing the unusual instances where internal and external splits are linked. Demonstrating that the splits are linked is not straightforward, and so I give four types of argument to establish that particular splits are indeed linked. The examples, from a range of languages, show subtle variations, and these differences allow us to see clearly the hierarchy of ways in which featural information is associated with lexemes.

We discuss the results of a video vignettes experiment that uncovers the variation of noun-classifier assignment in the possessive classifier system of six Oceanic languages. The results show that languages vary in their noun-classifier assignment, with some languages displaying relatively fixed assignment, similar to a grammatical gender system.

The term 'defectiveness' refers to gaps in inflectional paradigms — specifically, gaps which do not appear to follow from natural restrictions imposed by meaning or function. The Latin noun for 'change' is a textbook example: bizarrely, it lacks nominative and dative singular forms, and has no genitive plural. The fact that inflectional paradigms may have such anomalous gaps in them has been known since at least the days of the classical grammarians, but now as then, we understand little about them. And though the existence of defective paradigms is indisputable, few people could name more than a handful of examples. The project A Typology of Defectiveness, funded by the Arts and Humanities Research Council, completed in February 2009, has aimed to expand our empirical knowledge of this intriguing phenomenon and to clarify its significance for the study of language. This website hosts two complementary databases. The Typological Database examines the different types of defective paradigms according to various typological parameters. This database illustrates different types of defective paradigm according to various morphological and morphosyntactic parameters: a) Word class: what word class do the defective lexemes belong to? b) Morphosyntactic category: what morphosyntactic features characterize the missing forms? c) Paradigmatic correlation: can the gap(s) in the paradigm be described in terms of an easily definable morphological category (e.g. a word missing a particular morphological stem) or a morphosyntactic category (e.g. a verb missing its past tense or 1st person singular)? The 100-language Survey looks for plausible examples within a controlled sample, in order to gain a picture of how prevalent defectiveness actually is in the languages of the world.

Agreement systems often allow alternatives: This family has / have lost everything. Therefore typology requires a means for generalizing over them. Instances like plural have are frequently termed " semantic agreement " (vs. " syntactic agreement " for singular has), but this notion has proved difficult. The challenge is to encompass the full typological range of alternative agreements. These include the core instances: (i) hybrid nouns like family; and (ii) constructional mismatches, such as conjoined nominal phrases, but also less obvious phenomena: (iii) split hybrids where neither alternative is straightforwardly semantic, both appear related to form, and (iv) examples like Scandinavian " pancake sentences " , which stretch semantic agreement towards pragmatics. These different types are comparable in that (i) the alternatives are realized by the normal agreement forms; and (ii) they are subject to the Agreement Hierarchy. Hence they demand a common treatment. To achieve this, I first unpack the Agreement Hierarchy constraint into the agreement target positions and the directionality implied by " semantic agreement ". I show how the latter arises from mismatches between the agreement information available from different sources. Typically, in the core instances, the information from one source is more evidently semantic than from the second. But in other instances, this is less clear.

We start from the notion of ‘canonical’ inflection, and we adopt an inferential-realizational approach. We assume that we have already established the features and their values for a given system (while acknowledging that this may be a substantial analytic task). In a canonical system, feature values “should” multiply out so that all possible cells exist. Paradigms “should” be consistent, both internally (within the lexeme) and externally (across lexemes). Such a scheme would make perfect sense in functional terms: it provides maximal differentiation for minimal phonological material. However, real systems show great divergences from this idealization. A typology of divergences from the canonical scheme situates the types of morphological exceptionality, including: periphrasis, anti-periphrasis, defectiveness, overdifferentiation, suppletion, syncretism, heteroclisis and deponency. These types of exceptionality provide the basis for an investigation of higher order exceptionality, which results from interactions of these phenomena, where the exceptional phenomena target the same cells of the paradigm. While some examples are vanishingly rare, they are of great importance for establishing what is a possible word in human language, since they push the limits considerably beyond normal exceptionality.

This book aims to assess the nature of morphological complexity, and the properties that distinguish it from the complexity manifested in other components of language.

Lexicon schemas and their use are discussed in this paper from the perspective of lexicographers and field linguists. A variety of lexicon schemas have been developed, with goals ranging from computational lexicography (DATR) through archiving (LIFT, TEI) to standardization (LMF, FSR). A number of requirements for lexicon schemas are given. The lexicon schemas are introduced and compared to each other in terms of conversion and usability for this particular user group, using a common lexicon entry and providing examples for each schema under consideration. The formats are assessed and the final recommendation is given for the potential users, namely to request standard compliance from the developers of the tools used. This paper should foster a discussion between authors of standards, lexicographers and field linguists. Published in "Proceedings of the 6th International Language Resources and Evaluation Conference (LREC'08)"

Greenberg’s paper on universals (1963) contains an interesting set of generalizations relating to features. It is a good time to review the issues involved in establishing universals of features. These verge on the philosophical at one extreme, while at the other they concern the practical question of how we present and gloss examples. Various initiatives concerned with standardization, taken broadly, are under way, and it is important that they should be fully informed by the linguistic issues. There are two main areas to discuss: the Analysis problem and the Correspondence problem. The Analysis problem: for a given language, we need to be able to justify the postulation of any feature (such as number or case). Equally, for each feature in the language we need to be able to justify the set of values postulated (for example: singular, dual, paucal and plural; nominative, accusative and genitive). For some languages the analysis is trivially simple, in others it is exceptionally complex (for some there have been long-running debates). In this context, it is worth reviewing the work of the Set-theoretical School, given its undoubted relevance for typology. The difficulties posed by hybrids will be discussed; this leads naturally to typological hierarchies and the ‘Canonical’ approach in modern typology. The Correspondence problem: as typologists we need to be able to justify treating features and their values as comparable across languages. This is not straightforward, and yet a good deal of typology, including enterprises such as the World Atlas of Language Structures, depends upon it. The problem has a second, more subtle version. Even within a single language, features and their values do not necessarily line up consistently. In Bayso, the number system of nouns and verbs interact in a complex way. In Romanian, the genders of nouns and adjectives differ, and there are many more such examples. Here a typological perspective can inform the analysis of a single language and, of course, a typology which ignored these languages would be considerably impoverished. Features are an area where the concerns of the typologist meet those of computational linguists, formal linguists, fieldworkers, in fact linguists in many different guises. As we put increasing theoretical weight on features, it is important to review our assumptions and check our progress in understanding them.

Features are a central concept in linguistic analysis. They are the basic building blocks of linguistic units, such as words. For many linguists they offer the most revealing way to explore the nature of language. Familiar features are Number (singular, plural, dual, …), Person (1st, 2nd, 3rd) and Tense (present, past, …). Features have a major role in contemporary linguistics, from the most abstract theorizing to the most applied computational applications, yet little is firmly established about their status. They are used, but are little discussed and poorly understood. In this unique work, Corbett brings together two lines of research: how features vary between languages and how they work. As a result, the book is of great value to the broad range of perspectives of those who are interested in language.

We report a field study of the colour terms of the Bantu language Ndebele. The main purposes of the study were to describe the Ndebele colour term inventory, and to determine which colour terms were basic in order to test Berlin & Kay's theory of colour universals. A sample of school children and a sample of adults from Matabeleland performed a list task ('tell me as many colour terms as you know') in order to identify Ndebele colour terms and to establish the relative salience of the terms. The adults also took part in a low resolution colour mapping task which provided a preliminary estimate of the range of referents of each term. Our results show that Ndebele has four definite basic terms: terms for white, black, red and green with blue (grue). In addition, the children had a basic term for yellow, and as this term was taught formally in school, it may eventually become basic for the majority of the population. Both sets of basic terms are consistent with Berlin & Kay's theory.

This article provides a discussion on implicational hierarchies. It presents the examples of typological hierarchies and considers in turn syntactic, morphosyntactic, and lexical hierarchies. A well-known syntactic hierarchy is the Accessibility Hierarchy. The Agreement Hierarchy and the Animacy Hierarchy are the two well-established examples of morphosyntactic hierarchy. The Berlin and Kay Hierarchy is a famous typological hierarchy for lexis. Any proposed hierarchy must be justified by the range of data that it explains and the closeness of fit between the data and the claim made. The use of hierarchies for research on individual languages is described. The article finally deals with the extension of hierarchies and the relation of hierarchies to semantic maps.

Archi is spoken by about 1200 people in a remote mountain region in Daghestan. The language is characterised by remarkable phonetics, a very high degree of irregularity in all its inflecting word classes and by its morphological system, with extremely large paradigms. Archi culture is one of the most distinctive and best-preserved cultures of Daghestan.

The first book to examine these unusual gender systems in the canonical framework. Explores new data from a range of typologically diverse languages. Adopts a clear methodological approach in each chapter. Includes work from international experts in the field.

In this chapter, we investigate the variation in form, syntax and semantics of the plural words found across the Alor-Pantar languages. We study five AP languages: Western Pantar, Teiwa, Abui, Kamang and Wersing. We show that plural words in Alor-Pantar family are diachronically instable: although proto-Alor-Pantar had a plural number word *non, many AP languages have innovated new plural words. Plural words in these languages exhibit not only a wide variety of different syntactic properties but also variable semantics, thus likening them more to the range exhibited by affixal plural number than previously recognized.

The lexicon divides into parts of speech (or lexical categories), and there are cross-cutting regularities(features). These two dimensions of analysis take us a long way, but several phenomenaelude us. For these the term ‘split’ is used extensively (‘case split’, ‘split agreement’, and more),but in confusingly different ways. Yet there is a unifying notion here. I show that a split is an additionalpartition, whether in the part-of-speech inventory or in the feature system. On thisbase an elegant typology can be constructed, using minimal machinery. The typology starts fromfour external relations (government, agreement, selection, and anti-government), and it specifiesfour types of split within each (sixteen possibilities in all). This typology (i) highlights less familiarsplits, from diverse languages, and fits them into the larger picture; (ii) introduces a new relation,anti-government, and documents it; (iii) elucidates the complexities of multiple splits; and(iv) clarifies what exactly is split, which leads to a sharpening of our analyses and applies acrossdifferent traditions.*

Agreement is approached from the analytical decisions required for constructing a typological database. The Surrey Database of Agreement provides detailed, highly structured information on the agreement systems of fifteen genetically diverse languages. The range of material included and the criteria for inclusion are set out here. There is then detailed discussion of the difficult cases, in particular the dividing line between agreement markers and pronominal affixes. The criteria relevant to this distinction are in part drawn from the literature and in part new. The aim is that the criteria adopted should be fully clear, so that linguists of different persuasions can use the database for their varying purposes.

Typology in its modern form is connected with the search for universals. This works to the advantage of certain types of questions, those which allow a more or less coherent answer for any language. Phonology, syntax, and semantics are usually the starting point, and such topics as phonological inventories, word order, and the range of expressible semantic distinctions constitute the bulk of research. These also form the core questions of general linguistics, so this research emphasis is only to be expected. Conversely, one area that receives relatively little attention from typologists is morphology. This too is hardly surprising: of all the aspects of language, morphology is the most language-specific and hence least generalizable. Indeed, even the very presence of a meaningful morphological component is language-specific. © Walter de Gruyter 2007.

Proceedings of Corpus Linguistics 20003. University Centre for Computer Corpus Research on Language Technical Papers Vol 16

The aim of this paper is to demonstrate the existence of a highly restricted movement rule which operates after agreement rules. Syntactic and semantic evidence is given, and the data are taken from Russian, French and English.

Morphological features characterize variations in morphological form which are independent of syntactic context. They contrast with morphosyntactic features, which characterize variations in form correlated with different syntactic contexts. Morphological features account for formal variation across lexemes (inflectional class), as well as morphosyntactically incoherent alternations within the paradigm of a single lexeme. Such morphological features are not available to the syntax, as is made explicit in the principle of 'morphology-free syntax'. Building on work on stress patterns in Network Morphology and on stems in Paradigm Function Morphology, we take initial steps towards a typology of these morphological features. We identify four types: inflectional class features (affixal and prosodic), stem indexing features, syncretic index features and morphophonological features. Then we offer a first list of criteria for distinguishing them from morphosyntactic features (independently of the principle of morphology-free syntax). Finally we review the arguments demonstrating the need to recognize morphological features. © Springer Science+Business Media B.V. 2007.

Gender is a fascinating category, which has grown steadily in importance across the humanities and social sciences. The book centres on the core of the category within language. Each of the seven contributions provides an independent account of a key part of the topic, ranging from gender and sex, gender and culture, to typology, dialect variation and psycholinguistics. The authors pay attention to a broad range of languages, including English, Chukchi, Konso and Mohawk.

The Smith-Stark hierarchy, a version of the Animacy Hierarchy, offers a typology of the cross-linguistic availability of number. The hierarchy predicts that the availability of number is not arbitrary. For any language, if the expression of plural is available to a noun, it is available to any noun of a semantic category further to the left of the hierarchy. In this article we move one step further by showing that the structure of the hierarchy can be observed in a statistical model of number use in Russian. We also investigate three co-variates: plural preference, pluralia tantum and irregularity effects; these account for an item's behaviour being different than that solely expected from its animacy position.

In some Norwegian dialects, such as older Oslo dialect, the noun mamma ‘mother’ unexpectedly appears to be masculine. The Nordreisa dialect (Northern Norwegian) goes one step further. The word looks like it is masculine, but only in the definite form. This is an unusual “split” because gender mixture is normally based on number, not definiteness (but we find some few corroborative examples in other Norwegian dialects and different, but converging evidence on the Web). The Nordreisa example of mamma is unusual also because agreement targets are affected differently. The preference is for masculine agreement within the noun phrase, but for feminine agreement outside it. This is, therefore, an intriguing example since it combines a split based on definiteness with different gender requirements according to different agreement targets. On careful analysis, and given strict adherence to the classical, agreement-based definition of gender, the unusual behavior of mamma turns out to conform to the Agreement Hierarchy

In this scholarly volume, each of the living Slavonic languages are analyzed and described in depth, together with the two extinct languages--Old Church Slavonic and Polabian.

International Symposium 'the typology of argument structure and grammatical relations'11-14 may 2004, Kazan, Proceedings. Kazan state University

The Agreement Hierarchy consists of four principal target positions: attributive, predicate, relative pronoun and anaphoric personal pronoun. It constrains the distribution of alternative agreements, in that the likelihood of agreement with greater semantic justification increases monotonically as we move rightwards along the hierarchy. The Agreement Hierarchy covers a wide range of disparate data, and continues to figure regularly in work on theoretical syntax. Since the hierarchy was first proposed, typology has moved on. This means that to remain fit for the purposes for which it is currently used, the hierarchy needs an overhaul. The typology of agreement controllers is the area where the need is most urgent; this is therefore our focus. The canonical typology of controllers is shown to have two dimensions: lexeme to phrase, and local to extraneous (the latter involving honorific agreement, associative agreement, back agreement and “pancake sentences”). These two dimensions are amply illustrated. Finally, interactions between the different types of agreement controller are investigated, since these prove revealing for the typology. Besides making progress on the typology of agreement, the paper contributes to typology more generally, in incorporating insights from other typological disciplines.

Hybrid nouns, nouns which induce different agreements according to the target, have been described in various languages. The new question is why they exist at all. There is clear evidence that hybrids vary considerably in the agreement they control, even within a single language. It therefore seems logical to align this variability with lexical semantics, and this is convincing for some hybrids. But this motivation is hard to reconcile with the fact that some hybrids are hybrids only for part of their paradigm. These latter instances suggest that the underlying motivation for some hybrids is a form-meaning mismatch.

Nominal classification remains a fascinating topic but in order to make further progress we need greater clarity of definition and analysis. Taking a Canonical Typology approach, we use canonical gender as an ideal against which we can measure the actual gender systems we find in the languages of the world. Building on previous work on canonical morphosyntactic features, particularly on how they intersect with canonical parts of speech, we establish the distinctiveness of gender, reflected in the Canonical Gender Principle: In a canonical gender system, each noun has a single gender value. We develop three criteria associated with this principle, which together ensure that canonically a noun has exactly one gender value; we give examples of non-canonicity for each criterion, thus gradually building the typology. This is the essential groundwork for a comprehensive typology of nominal classification: the Canonical Typological approach allows us to tease apart clusterings of properties and to characterize individual properties with respect to a canonical ideal, rather than requiring us to treat the entire system as belonging to a single type. This approach is designed to facilitate comparisons of different noun classification systems across languages.

Person is required in an account of the syntax and the morphology of many languages, while others lack it. Between these two types are languages where person lacks unique morphological exponents (suggesting it is not a morphosyntactic feature) but interacts systematically with the expression of other features (suggesting it is a feature). In particular in a range of languages, notably in the Nakh-Daghestanian and Tucanoan families, the expression of gender and person are intertwined. The recurring pattern is that a default form in the gender system (inanimate or neuter) also serves for first and second person. After careful examination, possible analyses without a person feature become less attractive. While these genuinely difficult systems may still lead us to posit a morphosyntactic person feature, we must recognize that its status is intriguingly different from that which is normally found.

A native speaker of Russian 'knows' the gender of many thousands of nouns. While the literature on gender in Russian is considerable – as our bibliography shows – the basic question of how it is that Russians use gender correctly has been largely ignored. It is this question we shall try to answer. One hypothesis would be that a native speaker remembers the gender of each item individually; that is to say, each noun is specified for gender in his internalized lexicon. The opposite hypothesis would be that gender is never remembered – it can always be derived from other information about a noun (such as its meaning or phonological form); in other words, no noun is specified for gender in the lexicon. Between these extremes there are various hypotheses according to which the gender of many nouns would be derived by some sort of procedure, while that of various exceptions would be specified in the lexicon. We shall see that gender and morphological class are strongly interrelated, but we shall claim that the few investigators who have considered this relationship in any detail have postulated a type of dependency which creates more problems than it solves.

The Surrey Syncretism Database encodes information on inflectional syncretism in 30 genetically and geographically diverse languages, representing such morphosyntactic features as case, person, number and gender. Syncretism is defined as when some set of words fail to distinguish morphosyntactic feature values which we believe, based on language-internal criteria, to be underlyingly present (for example, in Latin, the dative and ablative cases may be distinct in some contexts but collapsed into a single form in others). For each language all instances of syncretism are recorded.

Contemporary linguistic theories distinguish the principal element of a phrase - the ‘head’ - from the subordinate elements it dominates. This pervasive grammatical concept has been used to describe and account for linguistic phenomena ranging from agreement and government to word order universals, but opinions differ widely on its precise definition. A key question is whether the head is not already identified by some other, more basic notion or interacting set of notions in linguistics. Heads in Grammatical Theory is the first book devoted to the subject. Providing a clear view of current research on heads, some of the foremost linguists in the field tackle the problems set by the assumptions of particular grammatical theories and offer insights which have relevance across theories. Questions considered include whether there is a theory-neutral definition of head, whether heads have cognitive reality, how to identify the head of a phrase, and whether there are any universal correlations between headedness and deletability.

Provides an introduction to the grammatical category of number surveying many of the world's languages.

The approach of Canonical Typology has proved fruitful for investigating a range of problems in syntax, inflectional morphology and most recently in phonology. It is therefore logical to take a canonical approach to derivational morphology. It provides a new perspective on some old issues, showing how previous key ideas fit together. The criteria proposed prove to have some degree of external justification. And from the point of view of canonical typology the results are particularly promising, since the criteria are interestingly different from those proposed in other domains.

Lexemes may have an internally consistent paradigm, or the paradigm may be split into segments. Splits may be ‘motivated’, that is they may correspond to morphosemantic, morphosyntactic2 or phonological specifications. Alternatively the split may lack such motivation, in which case we have a morphomic split, one which arguably increases the complexity of the system with no obvious corresponding return. We shall focus on the difference between these two types, so that we can recognise morphomic splits. There are some properties which the two types of split share: for instance, both motivated and morphomic splits can be viewed in terms of Wurzel’s Paradigm Structure Conditions (1989: 118), that is, there can be predictive relations within the segments; and both types can persist over long periods of time.3 But they are also interestingly different, which makes drawing the distinction valuable. It bears on the important notion that syntax is morphology-free. Our main question, then, is ‘how do morphomic splits differ from motivated splits?’

On-line proceedings of the 4th Mediterranean Morphology Meeting (MMM4), Catania, Sicilia 21-23 Sep 2003

It has been suggested that grammatical relations should be sufficient to determine agreement relations within the clausal domain. Three types of counter-example to this proposal are presented. Then evidence is presented which suggests that the rules for agreement require access to thematic roles and to communicative functions. In addition, they need to refer to surface case. While grammatical relations provide a useful part of a typology of agreement, they are far from sufficient.

Gender is a fascinating category, central and pervasive in some languages and totally absent in others. In this new, comprehensive account of gender systems, over 200 languages are discussed, from English and Russian to Archi and Chichewa. Detailed analysis of individual languages provides clear illustrations of specific types of system. Gender distinction is often based on sex; sometimes this is only one criterion and the gender of nouns depends on other factors (thus ‘house’ is masculine in Russian, feminine in French and neuter in Tamil). Some languages have comparable distinctions such as human/non-human, animate/inanimate, where sex is irrelevant. No other textbook surveys gender across this range of languages. Gender will be invaluable both for class use and as a reference resource for students and researchers in linguistics.

In a recent article, Wertz (1977) reviews the question of the number of genders in Polish. He considers previous answers to the question: three genders (Klemensiewicz 1965: 51), five (Mańczak 1956), six (Brooks and Nalibow (1970: 137) and himself proposes seven as the correct solution. It is interesting that an apparently straightforward question should be open to debate, and that there should be such a variety of answers. Naturally, different assumptions as to the nature of gender may produce different analyses; however, as gender is reflected in syntax at a superficial level it is relatively easy to test the adequacy of an analysis. I intend to show that even if we accept Wertz's assumptions, his seven-gender system is unable to handle the surface facts of agreement in Polish. More generally, the split between gender in the singular and gender in the plural, which Wertz and other scholars propose, is untenable.

The indigenous numerals of the AP languages, as well as the indigenous structures for arithmetic operations are currently under pressure from Indonesian, and will inevitably be replaced with Indonesian forms and structures. This chapter presents a documentary record of the forms and patterns currently in use to express numerals and arithmetic operations in the Alor-Pantar languages. We describe the structure of cardinal, ordinal and distributive numerals, and how operations of addition, subtraction, multiplication, division, and fractions are expressed.

Politeness has a major place in many languages, and is remarkably pervasive in some. Yet we rarely find respect as a morphosyntactic feature, alongside gender, person, number and case. I document this imbalance, and then ask why this is what we find. This paper was later published as a section of the book Features (2012). 338 pages. Cambridge University Press. ISBN 9781107661080. http://www.cambridge.org/gb/knowledge/isbn/item6832255/?site_locale=en_GB Also in Linguistik 2012, 51 (1). Available online at: http://www.linguistik-online.de/51_12/corbett_a.html

We set out to establish the 'basic' colour term inventory of Catalan, and to see if the inventory was consistent with Berlin and Kay's theory of colour universals. Pilot work had indicated that, like Russian, Catalan might have more than the maximum eleven 'permissible' basic colour terms allowed by Berlin and Kay's theory, and have more than one basic term for blue. A sample of adults and a sample of children performed a colour list task (in which they were asked to give as many colour terms as they knew) and a colour naming task. The list task provided a general indication of the likely basic colour terms, and the naming task indicated the referents of these terms. The results show that Catalan is best described as a standard stage seven Berlin and Kay language with just eleven basic colour terms. However, it has more than one salient term in the blue region and these correpond to the exceptional blue terms of Russian. Furthermore, whilst Catalan has a basic term for purple, its focus is unusually displaced from Berlin and Kay's 'universal' focus for purple, and this may be related to the way in which Catalan segments the blue region.

These studies examined naming and free-sorting behavior by informants speaking a wide range of languages, from both industrialized and traditional cultures. Groups of informants, whose color vocabularies varied from 5 to 12 basic terms, were given an unconstrained color grouping task to investigate whether there are systematic differences between cultures in grouping behavior that mirror linguistic differences and, if there are not, what underlying principles might explain any universal tendencies. Despite large differences in color vocabulary, there were substantial similarities in grouping behavior across language groups, and substantial within-language variation across informants. It seems that all informants group stimuli based on some criterion of perceptual similarity, but those with large color vocabularies are more likely to group stimuli in line with their basic color terms. The data are best accounted for by a hybrid system that combines a universal principle of grouping by similarity with culture-specific category salience.

The term 'defectiveness' refers to gaps in inflectional paradigms — specifically, gaps which do not appear to follow from natural restrictions imposed by meaning or function. The Latin noun for 'change' is a textbook example: bizarrely, it lacks nominative and dative singular forms, and has no genitive plural. The fact that inflectional paradigms may have such anomalous gaps in them has been known since at least the days of the classical grammarians, but now as then, we understand little about them. And though the existence of defective paradigms is indisputable, few people could name more than a handful of examples. The project A Typology of Defectiveness, funded by the Arts and Humanities Research Council, completed in February 2009, has aimed to expand our empirical knowledge of this intriguing phenomenon and to clarify its significance for the study of language. This website hosts two complementary databases. The Typological Database examines the different types of defective paradigms according to various typological parameters, while the 100-language Survey looks for plausible examples within a controlled sample, in order to gain a picture of how prevalent defectiveness actually is in the languages of the world.

Periphrasis straddles the border between two major linguistic components, morphology and syntax. It describes a situation where a grammatical meaning, such as a tense, which could be expected to be expressed morphologically within a word, is instead expressed by a syntactic phrase. Inclusion of syntactic phrases in morphological paradigms creates analytical and theoretical problems that have yet to be resolved by linguists, who have been hampered by the rather narrow range of data available for consideration and by a lack of adequate theoretical devices. This book addresses the challenge by broadening the range of phenomena under discussion and presenting new theoretical approaches to the problem of periphrasis. Part I takes four key languages from diverse families - Nakh-Daghestanian, Gunwinyguan (Australian), Uralic and Indo-European - as examples of languages in which periphrasis poses particular problems for current linguistic theories. Part II views periphrasis in different contexts, determining its place within the morphological and syntactic systems of the languages it is found in, its relations to other linguistic phenomena, and the typological variation represented by periphrastic constructions. Treating periphrasis as a morphological and syntactic phenomenon at the same time and applying the criteria worked out within the Canonical Typology approach allows linguists to view periphrasis as a family of phenomena within a typological space of syntactic constructions used to fulfil grammatical functions.

Inflectional morphology plays a paradoxical role in language. On the one hand it tells us useful things, for example that a noun is plural or a verb is in the past tense. On the other hand many languages get along perfectly well without it, so the baroquely ornamented forms we sometimes find come across as a gratuitous over-elaboration. This is especially apparent where the morphological structures operate at cross purposes to the general systems of meaning and function that govern a language, yielding inflection classes and arbitrarily configured paradigms. This is what we call morphological complexity. Manipulating the forms of words requires learning a whole new system of structures and relationships. This book confronts the typological challenge of characterising the wildly diverse sorts of morphological complexity we find in the languages of the world, offering both a unified descriptive framework and quantitative measures that can be applied to such heterogeneous systems.

© Oxford University Press, 2013.Features are standard currency in linguistics; they allow generalizations in syntax, and equally in morphology. Yet while features are heavily used, they are often taken for granted. It is therefore worth considering: the use of features (their logic, their place in different components), the substantive semantics of features, and the inventory of features.

Berlin & Kay's basic colour term framework claims that there is an ordering in the diachronic development of languages' colour systems. One generalisation is that primary colours, WHITE, BLACK, RED, YELLOW, GREEN, BLUE, are lexicalised before derived colours, which are perceptual blends, e.g. ORANGE is the blend of YELLOW and RED. The colour systems of Lower Sorbian and Upper Sorbian offer an important typological contribution. It is already known that primary colour space can contract upon the emergence of a basic derived term; our findings indicate that derived categories also shift as colour systems develop. Tsakhur offers corroborating evidence.

Features are central to all major theories of syntax and morphology. Yet it can be a non-trivial task to determine the inventory of features and their values for a given language, and in particular to determine whether to postulate one feature or two in the same semantico-syntactic domain. We illustrate this from tenseaspect-mood (TAM) in Kayardild, and adduce principles for deciding in general between one-feature and two-feature analyses, thereby contributing to the theory of feature systems and their typology. Kayardild shows striking inflectional complexities, investigated in two major studies (Evans 1995, Round 2013), and it proves particularly revealing for our topic. Both Evans and Round claimed that clauses in Kayardild have not one but two concurrent TAM features. While it is perfectly possible for a language to have two features of the same type, it is unusual. Accordingly, we establish general arguments which would justify postulating two features rather than one; we then apply these specifically to Kayardild TAM. Our finding is at variance with both Evans and Round; on all counts, the evidence which would motivate an analysis in terms of one TAM feature or two is either approximately balanced, or clearly favours an analysis with just one. Thus even when faced with highly complex language facts, we can apply a principled approach to the question of whether we are dealing with one feature or two, and this is encouraging for the many of us seeking a rigorous science of typology. We also find that Kayardild, which in many ways is excitingly exotic, is in this one corner of its grammar quite ordinary.

We specify a typology for the extreme of inflectional morphology, namely suppletion (as in go ~ went). This is an unusual enterprise within typology, and it requires a ‘canonical’ approach. That is, we define the canonical or best instance, through a set of converging criteria, and use this point in theoretical space to locate the various occurring types. Thus the criteria establish the dimensions along which we find the specific instances of suppletion, allowing us to calibrate examples out from the canonical. The criteria fall into two main areas, those internal to the lexeme and those external to it. Moreover, we find interactions with other morphological phenomena, and discuss four of them: syncretism, periphrasis, overdifferentiation and reduplication. These remarkable instances of suppletion, particularly when in interaction with other phenomena, extend the boundary of the notion ‘possible word’. Besides laying out the possibilities for the specific phenomenon of suppletion, we show how a canonical approach allows us to make progress in typology, even in the most challenging areas.

In descriptions of languages, we make use of morphosyntactic features such as gender, number or person. This paper shows that sometimes choosing the features and values to describe a language is not straightforward, and the decision of whether or not to use a particular feature requires careful consideration. Thus, when determining a language’s feature inventory, we should consider both why we posit a given feature, and how many values to posit for the feature. In our case study we look closely at the Daghestanian language Archi. It is usually assumed that languages have a person feature, but with Archi this is not self-evident. Archi (like some related languages) has no unique forms for agreement in person, and the standard descriptions of this language do not involve the feature person. However, the agreement patterns in Archi may be interpreted in favour of the presence of this feature, despite the absence of any phonologically distinct forms realising it. Thus, we claim that Archi does have the feature of person that had not been recognised for this language before. We also give a brief overview of the category of person in the languages of Daghestan.

Proceedings of the 5th International Conference on Computational Linguistics, Geneva

Agreement features introduce greater complexity into agreement systems than is generally recognized. They may determine the agreement domain (Dargi) and certain combinations of feature values can rule out particular sentence types (Tsakhur). Feature interactions show three levels of complexity: just the target may be involved (German), or a computation of controller feature values may be required (Slovene), or computation may involve a covert feature (Miya).

Often features are presented as clean, neat, simple. Indeed it is the contrast with the idiosyncrasies of lexical items which gives the intuitive justification for features. But reality is more complex. There are many instances where it is arguable whether we should postulate a feature (value), as with person in Archi. We must recognize that feature systems vary: (a) according to how well founded they are, and (b) in how they distribute across the lexicon. To analyse this difficult area, the penumbra of feature systems, I start from an idealized view, and then plot the deviations from that ideal. In other words, I take a ‘canonical’ approach. Having justified this approach in general terms, I propose a specific set of converging criteria for canonical features and values, concentrating on the genuine morphosyntactic features. In brief, the overarching principles are that a canonical morphosyntactic feature is constrained by simple rules of syntax (including the claim that syntax is ‘morphology-free’) and has robust formal marking. These give us a point in the theoretical space from which to calibrate the difficult instances which abound in feature systems. In accounts of particular features, various types of what we may call non-canonical behaviour have been pointed out: e.g., non-autonomous case values (Zaliznjak 1973), minor numbers, inquorate genders. We should ask whether these problems are feature-specific or whether they recur in the different morphosyntactic features. It turns out that, at the right level of abstraction, we find similar instances of non-canonicity with the different features. Let us concentrate on the criteria contributing to ‘robust formal marking’: Criterion 1: Canonical features and their values have dedicated forms. We find non-autonomous case values (violating criterion 1), in Classical Armenian, for instance (Baerman 2002); similarly we find non-autonomous gender values (as in Romanian). Criterion 2: Canonical features and their values are uniquely distinguishable across other logically compatible features and their values. Deviations give sub-genders (Serbian/Croatian/Bosnian), sub-cases (Russian) and sub-numbers (Biak). Criterion 3: Canonical features and their values are distinguished consistently across relevant parts of speech (word classes). In the easy examples, one part of speech has values which represent a collapsing of values available on another. More interesting are systems where combinations give additional values: combined gender systems (Mba), constructed number systems (Mele-Fila) and combined person (Maybrat). Criterion 4: Canonical features and their values are distinguished consistently across lexemes within relevant parts of speech. The basic deviation gives us a minor value: as in a minor number value (Bezhta), a minor case (Russian), a minor gender (Lelemi). This leads to the question of whether such deviations can co-exist. I give a striking example: the Russian second genitive exhibits all four types of non-canonical behaviour concurrently. Since it deviates from all four listed criteria this marks the extreme of the typological space. By investigating the penumbra of feature systems, including the possible and impossible interactions within the penumbra (which features are the context for the deviations of others), we put the theory of morphosyntactic features on a more realistic and hence firmer foundation.

The Grammatical Features Resources project is complementary to GOLD, and the motivation for offering the paper is to aim for useful interaction. The practical approach taken in GOLD has been to obtain a list of possible feature values from a selected source, with possible supplementation from other sources (Farrar, Lewis & Langendoen 2002). This sidesteps two major issues in constructing a feature/value inventory, namely the analysis problem and the correspondence problem. The proposed features site is intended to provide resources for addressing these substantial issues. Analysis requires us to state how we show that a particular language has, or has not, a given feature. The obvious ploy of requiring a phonological realization is shown to be inadequate for instances like person in Archi, where no single form demonstrates the existence of the feature and yet there are good arguments for person in this language. If the existence of a feature is demonstrable, we must then show how many values it has. Again requiring a phonological realization is too simple. There are instances in the literature of careful argumentation for difficult instances, notably the debate as to the number of case values in Russian (Zaliznjak 1973 and Comrie 1986). Establishing the values of number proved a long and difficult undertaking (given in particular the various references to quadrals in the literature, see Corbett 2000: 26–30). The site will offer two types of relevant information: first, the arguments which have been used to justify postulating features and values (with reference to the sources), and second, instances of challenging systems (which are typically those with most and sometimes with least values). Consider now the correspondence problem, starting with the simplest instance. French and Slovene both have masculine and feminine genders. Do they correspond? Yes, in the sense that nouns denoting females are typically assigned to the feminine gender. No, if we consider that French has two genders and Slovene three. But there are much more challenging cases. Romanian has two genders like French, if we look at agreement targets, but three like Slovene if we consider the nouns. The contribution here of the features resource site will be to provide outline typologies, so that labels such as ‘F’ or ‘N’ can be referred to a typology, clarifying the type of system within which they are functioning. The underlying philosophy, which is in tune with work on GOLD, is to offer useful tools for analysis and annotation.

The debate as to whether language influences cognition has been long standing but has yielded conflicting findings across domains such as colour and kinship categories. Fewer studies have investigated systems such as nominal classification (gender, classifiers) across different languages to examine the effects of linguistic categorisation on cognition. Effective categorisation needs to be informative to maximise communicative efficiency but also simple to minimise cognitive load. It therefore seems plausible to suggest that different systems of nominal classification have implications for the way speakers conceptualise relevant entities. A suite of seven experiments was designed to test this; here we focus on our card sorting experiment, which contains two sub-tasks — a free sort and a structured sort. Participants were 119 adults across six Oceanic languages from Vanuatu and New Caledonia, with classifier inventories ranging from two to 23. The results of the card sorting experiment reveal that classifiers appear to provide structure for cognition in tasks where they are explicit and salient. The free sort task did not incite categorisation through classifiers, arguably as it required subjective judgement, rather than explicit instruction. This was evident from our quantitative and qualitative analyses. Furthermore, the languages employing more extreme catego-risation systems displayed smaller variation in comparison to more moderate systems. Thus, systems that are more informative or more rigid appear to be more efficient. The study implies that the influence of language on cognition may vary across languages, and that not all nominal classification systems employ this optimal trade-off between simplicity and informa-tiveness. These novel data provide a new perspective on the origin and nature of nominal classification .

Some languages have both gender and classifiers, contrary to what was once believed possible. We use these interesting languages as a unique window onto nominal classification. They provide the impetus for a new typology, based on the degree of orthogonality of the semantic systems and the degree of difference of the forms realizing them. This nine-way typology integrates traditional gender, traditional classifiers and – importantly – the many recently attested phenomena lying between. Besides progress specifically in understanding nominal classification, our approach provides clarity on the wider theoretical issue of single versus concurrent featural systems.

A Dictionary of the Archi (Daghestanian) Language including word sounds and illustrations. Archi is spoken by about 1200 people in a remote mountain region in Daghestan. The language is characterised by remarkable phonetics, a very high degree of irregularity in all its inflecting word classes and by its morphological system, with extremely large paradigms. Archi culture is one of the most distinctive and best-preserved cultures of Daghestan

Contemporary linguistic theories distinguish the principal element of a phrase - the ‘head’ - from the subordinate elements it dominates. This pervasive grammatical concept has been used to describe and account for linguistic phenomena ranging from agreement and government to word order universals, but opinions differ widely on its precise definition. A key question is whether the head is not already identified by some other, more basic notion or interacting set of notions in linguistics. Heads in Grammatical Theory is the first book devoted to the subject. Providing a clear view of current research on heads, some of the foremost linguists in the field tackle the problems set by the assumptions of particular grammatical theories and offer insights which have relevance across theories. Questions considered include whether there is a theory-neutral definition of head, whether heads have cognitive reality, how to identify the head of a phrase, and whether there are any universal correlations between headedness and deletability.

This is the first book to present Canonical Typology, a framework for comparing constructions and categories across languages. The canonical method takes the criteria used to define particular categories or phenomena (eg negation, finiteness, possession) to create a multidimensional space in which language-specific instances can be placed. In this way, the issue of fit becomes a matter of greater or lesser proximity to a canonical ideal. Drawing on the expertise of world class scholars in the field, the book addresses the issue of cross-linguistic comparability, illustrates the range of areas - from morphosyntactic features to reported speech - to which linguists are currently applying this methodology, and explores to what degree the approach succeeds in discovering the elusive canon of linguistic phenomena.

SLS2006: The First Conference of the Slavic Linguistics Society, Bloomington IN.

This is a database that is freely available at The Surrey Morphology Group Website. http://www.smg.surrey.ac.uk/nws/ For more information also visit tp://www.surrey.ac.uk/englishandlanguages/research/smg/index.htm

This chapter discusses how deponency is to be differentiated from syncretism, which shows important similarities. It attempts to provide ‘intellectual housekeeping’, which puts order into the description of inflectional morphology. This allows for an analysis of the diversity of inflectional morphology by confronting it with an elegant order. It is revealed that there are types of lexeme whose interest and importance had not previously been fully recognized.

We need to bring together research into the diverse content of features in the world's languages with the discussion of their formal properties, and look for insights across sub‐discipline boundaries. This chapter offers summaries of all contributions and highlights areas of common ground between the different approaches. The selected perspectives represent major areas of linguistics where features are used.

Special issue of Transactions of the Philological Society 101, no.2

P[ossessive] A[djective]s in Slavonic, formed from nouns via suffixation, show unusual syntactic behavior. In Upper Sorbian, the form of attributive modifiers, relative pronouns, and personal pronouns can be controlled by the syntactic features of the noun underlying the PA. Control of attributive modifiers gives rise to phrases in which word structure and phrase structure do not match. The fact that the underlying noun is available for syntactic purposes suggests that PA formation is an inflectional process, while other factors (such as change of word-class membership) point just as clearly to a derivational process. It thus appears that any sharp differentiation between inflectional and derivational morphology must be abandoned. Data presented from all thirteen Slavonic languages, based on extensive work with native speakers, show that the control possibilities of the PA vary considerably. However, control of the attributive modifier is possible only if control of the relative pronoun is also possible, and that in turn only if control of the personal pronoun is possible. This result is subsumed under the constraints of the Agreement Hierarchy.

When examining Periphrasis we naturally analyse the state of the contributing elements and the interplay of syntactic and morphological factors. But if periphrastic forms are part of paradigms, we should also ask how periphrasis affects the notion ‘possible lexeme’. In particular, we can look at the ways in which periphrastic forms ‘split’ lexemes. This is a relatively new area. Since periphrasis is in large part the issue of whether we are dealing with a single lexeme or more than one, it is therefore worth investigating these splits more generally, first looking at ‘easier’ splits, and only then going on to the typology of splits created by periphrasis. I shall not go over the issue of the competition between synthetic and periphrastic forms, for which see, for instance, Stump (2002) and Kiparsky (2005); rather I shall concentrate on the shape of lexemes which are split by periphrasis.

We report an attempt to find more objective measures for identifying basic colour terms. We investigate the types of measure available, both linguistic and behavioural, and the statistical techniques for establishing the closeness of fit with the predictions derived from Berlin and Kay. This leads to an investigation of the interrelation between the measures; having examined consistency across investigators and across languages we establish that certain measures give considerably better results than others. While the indicators point in the same direction, supporting Berlin and Kay to varying degrees, different measures serve different functions.

We evaluate the design and implementation of the free-list experiment, which is a relatively easy method for exploring the membership of semantic categories (Weller & Romney 1988). Different factors that could affect the replicability and validity of the experiment are explored, and these are balanced with the need to work sensitively with speakers of endangered and minority language communities. By including aspects of a Participatory Research approach (van der Riet & Boettiger 2009), such as building rapport and respecting participants' knowledge, the experimenter can extend the free-list experiment to include wider discussions around the linguistic categories under study. We include a case study from our research on Oceanic possessive classifiers to show that a free-list experiment results in a wealth of data, and offers up opportunities for discussing and valorising different speakers' understanding of linguistic categories.

In a canonical inflectional paradigm, inflectional affixes mark distinctions in morphosyntactic value, while the lexical stem remains invariant. But stems are known to alternate too, constituting a system of inflectional marking operating according to parameters which typically differ from those of the affixal system, and so represent a distinct object of inquiry. Cross-linguistically, we still lack a comprehensive picture of what patterns of stem alternation are found, and hence the theoretical status of stem alternations remains unclear. We propose a typological framework for classifying stem alternations, basing it on the paradigm-internal relationship between the features marked by stem alternations versus those marked by affixes. Stem alternations may mark completely different features from the affixes (§2), or the same features (§3). Within the latter, the values may match (§3.1) – a rare situation – or be conflated (§3.2). Conflation in turn may involve natural semantic/morphosyntactic classes (§3.2.1), or phonological conditioning (§3.2.2), or be morphologically stipulated (§3.2.3). These patterns typically reveal stems’ continued allegiance to lexical as opposed to inflectional organizing principles.

Linguistics, and typology in particular, can have a bright future. We justify this optimism by discussing comparability from two angles. First, we take the opportunity presented by this special issue of to pause for a moment and make explicit some of the logical underpinnings of typological sciences, linguistics included, which we believe are worth reminding ourselves of. Second, we give a brief illustration of comparison, and particularly measurement, within modern typology.

The papers in this volume can be grouped into two broad, overlapping classes: those dealing primarily with case and those dealing primarily with grammatical relations. With regard to case, topics include descriptions of the case systems of two Caucasian languages, the problems of determining how many cases Russian has and whether Hungarian has a case system at all, the issue of case-combining, the retention of the dative in Swedish dialects, and genitive objects in the languages of Europe. With regard to grammatical relations, topics include the order of obliques in OV and VO languages, the effects of the referential hierarchy on the distribution of grammatical relations, the problem of whether the passive requires a subject category, the relation between subjecthood and definiteness, and the issue of how the loss of case and aspectual systems triggers the use of compensatory mechanisms in heritage Russian.

Gender in Chichewa is described as a complete system. First the basic data on gender agreement are presented and it is shown how the available agreement markers correlate with the noun genders (and how the system has changed in the recent past). There follows a discussion of interesting phenomena which do not fit easily into the main gender system. Next structures involving conjoined noun phrases headed by nouns from various genders are analysed in detail. The rules required to account for the Chichewa system prove particularly complex; rules proposed for other Bantu languages do not cover all the Chichewa facts. The data are important for comparative work within Bantu and for typological claims which go beyond.

If we consider the widely varying approaches to languages, we find one thing that is shared by almost all: namely the use of features. They have a central place in theoretical syntax and morphology, and are the subject of major typological generalizations. Although features underpin a good deal of what we do in linguistics, they have been neglected: they are used in inconsistent ways, without sufficient attention to the logic of their use and the variety of their meanings. The course will therefore consider why features are so important in linguistics, and set out the different types of feature. We then consider the basic and challenging issue of how we establish the features and values of a particular language. We then have to ask whether and how we can compare features across languages. We whall analyse particularly the genuinely morphosyntactic features (number, gender, person, case, and in rare instances definiteness and respect), since these are in many ways the most interesting. We shall see that they do indeed offer interesting typological patterns, while also displaying remarkable diversity.

The verbal paradigms of the Daghestanian language Archi are justly famous for their impressive size. I argue, however, that there is a more difficult problem lodged within a small and apparently simple part of the paradigm. It concerns the expression of gender and number, in their interaction with person. I present information on the large scale of the paradigm briefly, and then outline the problem of person. The need, or lack of it, for a person feature in Archi has been discussed elsewhere, so that here I can simply summarize the argument.If the need for a person feature is accepted, it follows that the paradigm has an unusual shape. This paradigm is genuinely difficult, as I demonstrate in the main part of the paper."

Studies have challenged the assumption that different types of word-final s in English are homophonous. On the one hand, affixal (e.g. laps) and non-affixal s (e.g. lapse) differ in their duration; on the other hand, variation exists across several types of affixal s (e.g. between the plural (cars) and genitive plural (cars')). This line of research was recently expanded in a study in which an interesting side effect appeared: the s was longer if followed by a past tense verb (e.g. The pods/odds eventually dropped), in comparison to a following present tense verb (e.g. The old screens/jeans obviously need replacing.). Put differently, the s became longer in the absence of overt morphosyntactic agreement, where it was mostly the sole plurality marker in the sentence. The objective of the present article is to examine whether this effect can be replicated in a more controlled setting. Having considered a large number of potential confounding variables in a reading experiment, we found an effect in the expected direction, one that is compatible with the literature on the impact that predictability has on duration. We interpret this finding against the background of the role of fine acoustic detail in language.

Categorization retains its key importance in research on human cognition. It is an intellectual area where all disciplines devoted to human cognition – psychology, philosophy, anthropology, and linguistics – intersect. In language, categorization is not only a central part of lexical structure but is also salient in systems of nominal classification, notably gender and classifiers. Recent years have seen great progress in the description and analysis of nominal classification systems, so that we are now in a position to offer an account of such systems which brings cognition and typology together, providing the essential parameters for the calibration of experiments for investigating cognition. To this end, we establish the extremes of nominal classification systems, from the surprisingly simple to the surprisingly complex. We analyse the two essential components of nominal classification systems: (i) assignment, i.e. the principles (semantic or formal) which govern category assignment and (ii) exponence, i.e. the morphological means by which systems of nominal classification are expressed. We discuss extreme configurations of assignment and exponence in individual languages and extreme multiple pairings of assignment and exponence in languages with two or even more concurrent classification systems.