Gregory Stump

Position classes and morphological theory

Two generations of linguistics students have cut their teeth on the morphological analysis problem in (1):

Morphological and syntactic paradigms: Arguments for a theory of paradigm linkage

A paradigm is a set of cells; each such cell is the pairing of a word form with the set of morphosyntactic properties which that word form realizes. Thus, the paradigm of the Latin noun amicus ‘friend’ is the set of cells in (1).

The Interpretation of Frequency Adjectives

ConclusionWhat I have attempted to show in the foregoing is that: (1) For any frequency adjective f, there is an element of meaning common to both the adverbial and the generic usage of f; this is a function f′ from propositions to truth-values such that f′(Φ)′ is true at an interval i iff Φ′ is true at subintervals of i distributed through i in a certain way. (2) In an adverbial use of f, f′ functions like the corresponding frequency adverb. (3) In a generic use of f, f′ quantifies the times intervals which a specified object is realized by some stage. (4) In their adverbial usage, frequency adjectives are not regular attributive adjectives at the level at which interpretation takes place, but are perhaps determiners. And (5) in their generic usage, frequency adjectives are adjectives at the level at which interpretation takes place.More generally, I hope to have suggested something of the range of uses to which frequency is put in natural language. That we effortlessly grasp what is meants by Monday, the Columbus Dispatch, autumn, the Today Show, and so on, demonstrates our fluency in interpreting certain sequences of objects, events, or states of affairs as sequences of values of some function cycling on a time-axis. Much more must ultimately be told about the psychological and anthropological significance of frequency.

How Peculiar is Evaluative Morphology

Many languages possess morphological rules which serve to express diminution or augmentation, endearment or contempt; examples are the Breton rule relating potr ‘boy’ to potrig ‘little boy’, the Shona rule relating chibikiso ‘cooking tool’ to zichibikiso ‘huge cooking tool’ and the Italian rule relating poeta ‘poet’ to poetastro ‘bad poet’. Because of the possibility of interpreting diminution and augmentation in affective rather than purely objective terms (Wierzbicka, 1980: 53ff.; Szymanek, 1988: 106ff.), morphological expressions of diminution or augmentation are not always discrete from those of endearment or contempt; that is, diminutives and augmentatives are frequently used as expressions of endearment (such as Italian sorella ‘sister’ → sorellina ‘dear little sister’, donna ‘woman’ → donnotta ‘fine, stout woman’) or disdain (Italian uomo ‘man’ → uomicciuolo ‘contemptible little man’, donna → donnona ‘overgrown girl’).

Template morphology and inflectional morphology

In recent years, attention has been drawn to the possibility that morphology might be of two radically different types; these two types have been labelled — infelicitously, in my view — as layered morphology and template morphology. Intuitively, a morphological expression exhibits template morphology if the ordering of its affixes follows not from the properties of the individual affixes themselves (or of the rules introducing them) but from an independent stipulation about the ordering of these affixes (or rules). By contrast, a morphological expression exhibits layered morphology if the ordering of its affixes is simply an effect of the properties of the individual affixes themselves (or of the rules introducing them). Here, I shall address two fundamental questions about the distinction between template and layered morphology, namely (1) and (2): (1) How does the distinction between template and layered morphology relate to the distinction between inflectional and derivational morphology? (2) What form should templates take in an adequate theory of morphology?

On the theoretical status of position class restrictions on inflectional affixes

In heavily inflected languages, a word’s inflectional affixes are often required to appear in a fixed sequence; in such cases, the affixes can be classified according to the position(s) which they may occupy in that sequence. Consider, for instance, the partial inflectional paradigm of the Swahili verb ‘to want’ in (1). In this partial paradigm, three distinct prefix position classes can be distinguished, as in (2): the slot nearest the verb root is occupied by a class of prefixes which mark tense, slot II is occupied by a class of prefixes marking person/number agreement, and slot III is occupied by a class of prefixes marking negative polarity.1

A Note on Breton Pluralization and the Elsewhere Condition

Anderson (1986) has argued that Breton ‘double plurals’, though apparently disconfirming the Elsewhere Condition, can be brought into conformity with it if they are viewed as deriving from basic collective nouns rather than from plural nouns. This solution is here shown to be unworkable: on the one hand, certain double plurals derive from forms with transparently plural morphology; moreover double plurals counterexemplify the Elsewhere Condition even if they are assumed to derive from basic collectives, since the latter are not distinct in their morphosyntactic feature content from ordinary plurals. Plural diminutives present similar difficulties, since their formation requires the successive application of two rules whose application is predicted by the Elsewhere Condition to be disjunctive. Besides suggesting that the Elsewhere Condition cannot be maintained in its strongest form, the Breton evidence raises questions about the existence of a strict division between inflectional and derivational morphology.

Paradigm Function Morphology

Paradigm function morphology is an inferential-realizational theory of inflection in which the definition of a languages inflectional morphology makes essential reference to the notion of paradigms. In formal terms, this theory equates the definition of a languages inflectional system with the definition of its paradigm function, a function from the grammatical cells in its paradigms to their phonological realizations. Paradigm function morphology affords simple accounts of a number of morphological phenomena, including morphological defaults, head-marking, complex position-class systems, syncretism, heteroclisis, deponency, and periphrasis.

Generating Hebrew Verb Morphology by Default Inheritance Hierarchies

We apply default inheritance hierarchies to generating the morphology of Hebrew verbs. Instead of lexically listing each of a word forms various parts, this strategy represents inflectional exponents as markings associated with the application of rules by which complex word forms are deduced from simpler roots or stems. The high degree of similarity among verbs of different binyanim allows us to formulate general rules; these general rules are, however, sometimes overridden by binyan-specific rules. Similarly, a verbs form within a particular binyan is determined both by default rules and by overriding rules specific to individual verbs. Our result is a concise set of rules defining the morphology of all strong verbs in all binyanim. We express these rules in KATR, both a formalism for default inheritance hierarchies and associated software for computing the forms specified by those rules. As we describe the rules, we point out general strategies for expressing morphology in KATR and we discuss KATRs advantages over ordinary DATR for the representation of morphological systems.

Some criticisms of Carstairs-McCarthy’s conclusions

In his paper “Affixes, stems and allomorphic conditioning in Paradigm Function Morphology”,AndrewCarstairs-McCarthy (hereafterC-M) reacts to thoseparts of Inflectional Morphology (IM) that relate to his own research interests; here, I show that the central assertions in his paper rest on empirically indefensible assumptions and on faulty argumentation. The main points in C-M’s paper are that, contrary to the assumptions of IM: (a) concatenative andnon-concatenative inflectionplaydistinct roles in thearchitectureof a language’smorphology; (b) an affix can be related to a setσ ofmorphosyntactic properties in either of twoways: it can be an exponent of σ (i.e σ can specify its content), or it may require some accompanying piece ofmorphology to realise σ (i.e. σ can specify its context); (c) the association of a morphomic stemwith particular cells in a lexeme’s paradigm is not effected by multiple rules of exponence; and (d) a theory incorporating property co-occurrence restrictions is nomore restrictive than one incorporating disjunctive rule ordering. C-M’s argument in support of (a) is that the No Blur Principle (CameronFaulkner and C-M 2000) entails a fundamental distinction between concatenative and non-concatenative inflection; his argument in support of (b) is that the Peripherality Constraint (Carstairs 1987) entails a fundamental distinction between content properties and context properties. Thus, he argues that the assumptions of IM are incompatible with the No Blur Principle and the PeripheralityConstraint.But as I show in sections1and2, themere factof this incompatibility is not a convincing argument against the assumptions in IM, since neither theNoBlur Principle nor the Peripherality Constraint can be validlymaintained. In support of conclusion (c),C-Margues that the cells in a lexeme’sparadigmwith which amorphomic stem is assocated are stipulated enbloc; as I show in section 3, this claim is empirically unsupported. In support of conclusion (d), C-M argues that by recurring to property co-occurrence restrictions, virtually any instance of disjunctive ruleordering canbemimicked; as I show in section4, this claim is false. Though my arguments here are intended to point out the errors in C-M’s paper, I hope that theywill bemore generally seen as a reminder that in the study of human language, the need to subject one’s assumptions to a high standard of empirical verifiability is unrelenting.