Illocutionary force, on the other hand, is 'conventional, in the sense that at least it could be made explicit by the performative formula. Thus we can say I argue that or I warn you that but we cannot say I convince you that or I alarm you that' (103a; cf. also 120–21). Moreover, it is possible (73–6) to establish a kind of scale of illocutionary equivocation, ranging from explicitly performative sentences like I permit you to shut it—through sentences with some relatively unambiguous device for indicating IF, such as mood (cf. Shut it, if you like)—to highly ambiguous sentences where the IF is shown only by the non-linguistic circumstances of the utterance, notably simple declaratives. And finally, it is no harder to decide on the IF's of particular utterances than it is to decide on their sense and reference: 'A judge should be able to decide, by hearing what was said, what locutionary and illocutionary acts were performed, but not what perlocutionary acts were achieved' (Austin, 121; cf. also 114–15).

This article discusses two aspects of definite and indefinite NP's in the grammar of Spanish: SPECIFICITY, which is formally marked by the mood of restrictive relative clauses; and EXISTENTIAL IMPORT, which derives from the linguistic environment by principles connected with presuppositions in general. Specificity and existential import belong to the grammar of Spanish, not to the realm of pragmatics. In Spanish, Donnellan's referential descriptions constitute a sub-set of specific NP's (definite specific NP's in the singular with an existential presupposition determined by the linguistic environment). Attributive descriptions, on the other hand, are singular, definite, nonspecific NP's having an existential presupposition.

In Spanish non-sentence constructions, beginning elements establish a datum of reference to what is already known by carrying meaning within the confines of conventionalized order-class paradigms. Following elements, on the other hand, narrow the possibilities of syntagmatic combination to give final, new definition to the expression.

At the beginning of constructions, the potential paradigms that occur first are smaller in extension, and thus more contrastive; paradigms nearer the center are larger, and more complex. The value of relative complexity or contrast of each such paradigm is closely connected to its characterizing semantic tag.

Following elements possess syntagmatic contrast similar to the contrast perceived in initial paradigms; preceding elements, which are non-contrastive, are perceived as syntagmatically complex. Phrases and clauses, by nature complex, are placed at the end of constructions, allowing the possible modulation or neutralization of end contrast.

The theoretical observation that certain types of pronominal differentiation are functionally related to certain types of transformational rules leads to a loose concept of a ‘functional component’ (independent of formal relationships), within which these two aspects of grammar play complementary roles. This complementarity is matched by an empirically observed inverse relationship in the functional values of the two from one language to another. By looking at the functional component as a whole (i.e. at a system rather than at individual elements), we arrive at an interesting new kind of universal.

English has two classes of modal deference expressions which may be superordinate to performative verbs: one asks permission from the addressee to perform the act; the other expresses the speaker's frame of mind in performing the illocutionary act. These occur with performative verbs in certain restricted ways because of their associated senses and presuppositions, but they have no illocutionary force of their own. Thus an utterance cannot constitute an intended illocutionary act if there is a conflict in the speaker's pragmatic presupposition between the addressee-authority and the speaker-authority.

Kiparsky 1973 presents evidence that the treatment of disjunctive ordering in The sound pattern of English is inadequate, and proposes an alternative set of general principles collectively referred to as the Elsewhere Condition. The purpose of the present article is to evaluate the adequacy of the Elsewhere Condition on the basis of (1) a review of the evidence given in its favor, and (2) an examination of some of its implications. It is concluded that there is little evidence to warrant the incorporation of the Elsewhere Condition into linguistic theory, and good reason to doubt its validity.

Kiparsky 1973b proposes a number of constraints on global rules. All his examples are of the type where derived strings become separated, in terms of their behavior, from otherwise identical underlying strings. This paper will argue for cases of global rules that do not meet Kiparsky's conditions. In fact, the two cases discussed violate Kiparsky's conditions by applying to underlying rather than derived strings. The inescapable conclusion is that Kiparsky's constraints are not valid for grammars.

Indo-European offers a wide variety of infinitive formations, many of which are shared by several of the dialects, and all of which appear to reflect earlier verbal nouns. The literature includes considerable discussion on the etymological origins of individual infinitive endings, but little on intermediate developments which may account for the disparate systems of the historical languages. This paper suggests that at least two separate classes of infinitives are to be noted among the infinitive types which occur cross-dialectally, and that these two classes reflect chronologically distinct occurrences in the history of IE infinitival development. This distinction is most clearly demonstrable in Vedic, but parallel evidence exists in several other dialects.

This investigation shows that a non-English portion of the universal initial-cluster hierarchy is cognitively represented in monolingual English-speaking children and adolescents. Since the representation is stronger in adolescents, the universal structure of the hierarchy may be due to a universal acquisition process rather than to innate linguistic content.

+CCVC+ syllables were developed representing seven types of initial non-English consonant clusters, from a universal hierarchical ordering of the type IF +WX IS AN INITIAL CONSONANT CLUSTER IN A LANGUAGE, THEN + YZ IS ALSO AN INITIAL CONSONANT CLUSTER IN THAT LANGUAGE (Greenberg 1965), for which markedness matrices have been developed (Cairns 1969). Children (N = 40) and adolescents (N = 40) listened to a tape and read a corresponding written representation of twenty-four pairs of syllables which differed only in their initial cluster. Subjects were to decide which syllable in each pair probably occurs in more languages. Both adolescents and children were able to reconstruct that portion of the phonological hierarchy which was tested, although adolescents did so significantly better than children. The effect of psychological distance from English (Greenberg & Jenkins 1964) was demonstrated on the same subjects with separate materials, but did not account for the hierarchy effects, since the syllables were controlled for that variable.

Finally, responses of subjects confirmed a theoretical error in previous works, suggesting an alternate arrangement of part of the phonological hierarchy.