Search

This paper studies the so-called strict subcategorization rules, and the theory associated with them, in the transformational grammar of. Erigl·ish as proposed by Noarn Chomsky in his Aspects. The syntactic component of English transformational grammar consists of two mutually ordered parts, viz., the base and the transformational subcomponents. The initial part of the base are the so-called categorial rules, which are of almost exclusive interest to us here. Their primary task is to generate what are usually called basic sentence patterns, and will here, with Chomsky (Aspects, p.ll3), be designated with the expression, frames of category symbols.- The rules of the transformational subcomponent modify, in various ways, the frames generated by the base. For several reasons - one of them being that the correct work of the transformational subcomponent quite often depends on the kind of lexical items with which the syntactic positions in the frames of category symbols have been filled, the lexical items must be introduced from the lexicon into the empty positions in the frames before the rules of the transformational subcomponent can be allowed to modify the frames.

The present paper is a list of Modern Icelandic nouns, adjectives, and adverbs, analysed into their respective stems and endings; the declension of the suffixed definite article is also included. Under each item it is stated which rules, if any, apply in the derivation of its grammatical forms. The following items of the list should be consulted for new phonological rules: (3), (11), (12), and (133). A grammatical innovation has been implemented in the list, namely the so-called REPLACING ENDINGS. These are not added after the last segment of the stem, as endings usually are, but replace the last segment(s) of the stem. More is said on replacing endings in the Introduction.

The paper deals with the phonology of modern Icelandic preaspiration. The rule that produces the non-lexicalised cases of preaspiration before what is written as p, t, k + l, n is different for hardmadi and linmeli. The hard medi role generates preaspiration before an aspirated plosive + /1, n/. (The aspiration of the plosive is obliterated by a later deaspiration rule.) The linmaeli rule generates preaspiration before plosive + /l, n/ just in case there is at least one such form in the inflexional paradigm of the word-form to which the rule is to apply containing a long vowel immediately followed by a plosive (where the vowel and the plosive are those mentioned in the structural description of the preaspiration rule).

Before the book under review, Magnús Pétursson (=MP) had published his doctoral dissertation about the articulatory phonetics of Icelandic (1974, in French) , a book on general and Icelandic phonetics, and another on general and Icelandic phonology (1976, 1978, both in Icelandic) in addition to a number of papers connected with phonetics and phonology: MP has now to be commended for having tackled a broader subject: Isländisch is an attempt at presenting an overall view of Iceland and Icelandic, especially the latter, to the German-reading public.

The present paper is a list of modern Icelandic verbs and pronouns analysed into their respective stems and endings. Under each item it is stated which phonological rules, if any, apply in the derivations of its grammatical forms. The following items of the list should be consulted for new phonological rules: (3), (6), (31), (37), and (49); see also the Appendix. A grammatical innovation has been implemented in the list, namely the socalled replacing endings; see the Introduction.

This paper addresses the problem of defining the notion of mixture in languages and dialects as used by Jan Baudouin de Courtenay (1845 - 1929). Focussing on the types of interference which were formulated on the basis of observations of the dialects of the Slovene language (5) , the paper deals primarily with those Baudouin's theoretical positions (4), observations and deductions (5.1) which are today part of a theory of linguistic interference. The article is an English version of a paper, presented at the Eighth International Congress of Slavists which was held in Zagreb from the 3rd to the 9th September 1978.

It has long been suggested that the Hittite verb da-ah-hi 'I take' is to be connected with the Indo-European root *dō- ' to give'. (See Friedrich, 1952, 201; Kronasser, 1966, 533-534; Kammenhuber, 1969, 235-236). The pur­pose of this paper is to support this suggestion but in a slightly different way than it has been supported in the past.

In this paper we shall focus our attention. on several aspects of the ongm and nature of homonymy, drawing examples from Russian and Serbo-Croatian and other Slavic languages. We shall first view this from a diachronic perspective noting some of the reasons for the appearance of homonyms. Then we shall examine the phenomenon at the synchronic level employing a variant of Serbo-Croatian in which important phonetic changes now in pro­ cess have given rise to new homophones. We may begin by turning to the ancestor of both Russian and Serbo-Croatian, Common Slavic. Homonymy appears to be a linguistically universal phenomenon, and it is safe to assu­ me that there were homonyms in Common Slavic. One pair of such words was undoubtedly the infinitive pasti 'to graze' (pasq) and ' to fall' (padq) which has its reflexes in most of the modern Slavic languages, e. g. Serbo-Croatian pásti and pȁsti or Russian pastí and past' and Ukrainian pásty.

He who reads carefully Cicero'sDe Legibus and Martin Luther King, Jr.'s Stride Toward Freedom is surprised to find, mutatis mutandis, on how many points these two great men agree. The historical circumstances are different, but the essential ideas are very similar. The purpose of this paper is to show on what precisely they agree and on what they differ.

In this paper is presented a new hypothesis regarding the development of the so-called "endingless locative" construction of Indo-European. It is argued that locative formations in -ē, and * ī came to appear in the , o-stem nouns through the monophthongization of word-final *-oi and *-ei before word-initial consonants. These formations were analogically extended to the other stem-classes, with the suffix*-ē, frequently hypercharacterized by the locative particles *-i or *-u, becoming the primary marker of the locative function in the i- and u-stems. On the basis of such apparently "endingless" locative constructions, old consonant-stem nominative (-accusative) formations in were reinterpreted as locatives as well.

This paper discusses the velar pronunciation of the root final segments in the plural of the Icelandic nominalized present participles, e.g. in leikend-, the plural of leikandi "actor". Since k(k)and g(g) are pronounced palatal before e in other old words(e.g. in akkeri "anchor"), the velar pronunciation of leikend- and similar words requires an expla­ nation. I see in the non-application of the palatalisation rule (while that rule operated before e; it is by no means certain that it still does) to leikend- and similar words the effect of Kiparsky's phonological universal ("Neutralization processes apply only to derived inputs") as modified in Orešnik 1979 (substrings crucially involving word-derivational morpheme boundaries are not derived inputs). Thus the /k+e/ of leikend- is not a derived input, and the palatal1sation rule, a neutralisation process, does not apply to it.

In this paper it is suggested that the original form of the Indo­ European third person plural verbal suffix was *-(e/o)N and that the nasal element which appears in this suffix was or ginally a deictic particle with 'there and then' signification.

The paper describes the formal and semantic properties of the mood and aspect categories of the Adamawa language, Karang. Three inherent aspect verb classes are established--events, processes, and states--on the basis of semantic and morphological distinctions. A fundamental opposition of the mood-aspect system is between factive and non-factive moods, which distinguish actual and potential situations. Non-factive mood is formally indicated by a high tone and subdivides into the categories subjunctive, predictive, and nonpredictive. Verbo-nominals are marked as non-factive. The formal categories of aspect are progressive, habitual, perfect, and nonperfect. When inherent and formal aspect categories with semantically contradictory components are combined, inherent aspect is overriden. The perfective meaning of the perfect category also overrides the imperfective meaning of the progressive.

Downstep in the vast majority of cases can be traced to the influence of a low tone [Hyman 1979]. This paper discusses the case of Supyire, a three tone language with downstep unrelated to low tone. Sequences of high tones are automatically downstepped. Confusion with mid tone is in most cases avoided by a rule which raises mid to high tone; the two rules together creating a "flip-flop" effect. Supyire also has automatic downdrift of mid and high tones following a low tone. It thus presents an unusual combination of downstep and downdrift in the same language, but unrelated to each other.

This paper will concentrate on a type of verb complement structure referred to as the HTS construction in Yoruba, essentially to show four points: a. that it is an older form of the verb-infinitive phrase in Yoruba, out of which the infinitival gerunds, the a-ti complements, and (by a historical change) the serial verbal constructions originate; b. that the HTS construction is different from (and should not be confused with) the infinitival gerunds, since they are demonstrably distinct constructions; the infinitival gerunds which do not have HTS equivalents number in the hundreds, far too many to be exceptions; c. that the HTS constructions show Yoruba to be related to neighboring Kwa languages like Igbo, Itsekiri and Efik. Furthermore, the serial verb constructions will be shown to be far more productive and economical than the HTS constructions, the infinitival gerunds, and the a-ti complements; hence the former's gradual replacement of the latter, probably to suggest a direction of change (from relative complexity to relative simplicity and productivity); and lastly d. that it will become necessary to modify a portion of the orthography of the language to accomodate these new facts.

This paper addresses the general question of genetic vs. nongenetic language development, in the context of a structural and historical discussion of Ma'a (Mbugu), a language with Cushitic basic vocabulary that is spoken in Tanzania. The grammatical structure of Ma'a is compared to characteristic Cushitic and Bantu structures. The conclusion that emerges from this comparison is that Ma'a probably does not have enough Cushitic grammar to qualify as a Cushitic language in the full genetic sense; and if it does not, its origin must be nongenetic. The final section of the paper seeks to determine the particular route of nongenetic development that Ma'a has followed, using the direct evidence of published comments about its speakers' history and the indirect evidence of comparison with other languages whose origin is nongenetic or, like Ma'a, on the borderline between genetic and nongenetic.

This paper uses the concept of discourse cohesion as a basis for looking at certain sentence emphasis patterns in Pulaar (Fula). In particular, four patterns of focus are examined, first syntactically and then pragmatically. Focus in Pulaar can be divided according to clefting and pseudo-clefting patterns for subject and non-subject focus. These four patterns are discussed in depth. Then, I suggest that the idea of discourse connectedness provides a model for explaining the choice of a focus pattern in a given environment. Examples of this are drawn from a transcription of the story of Gelaajo Ham Bod'eejo, performed by Mammadu Non Gii se, a griot from northern Senegal.

This paper examines those complex phonetic elements in Bura (and related languages such as Margi) which have been described by earlier scholars as unitary segments having a labial-alveolar double articulation. Several kinds of observations at the phonetic level show that this description is incorrect. Instead these segments are shown to be composed of a phonetic sequence of a labial articulation and an alveolar one. The analysis of the "labial-alveolars" as a sequence of phonological elements is also shown to be appropriate. If they are viewed as unitary segments represented by a single matrix of features there are problems with selecting correct feature values. If they are represented as two feature matrices linked to a single segmental position an invalid distinction is introduced between this representation and the same two matrices linked to separate segmental positions. On the other hand, representation as two independent segments accounts for vowel length adjustments and for resyllabification. Although unusual syllable-initial consonant clusters are posited in this analysis, these clusters are heterosyllabic when an appropriate environment occurs and violations of the universal syllable canon are minimized.

Ergativity would seem to be non-existant or at least quite rare in Africa. This lack, however, may be related to another continent-wide areal phenomenon: there is a paucity of morphological NP case marking according to either ergative or accusative typologies. It is thus possible that other more subtle attributes of the ergative organization of syntax are what should be sought in Africa. For example, in the Mande languages, as also in Celtic, phonological decay has produced a series of word initial consonant alternations. In Celtic these have come to function as part of a nominative-accusative case marking strategy. The situation is quite similar in Mande, but as this paper details for Lorna, the noun case system is ergative-absolute. And, accordingly, the pronoun system has active-stative characteristics.

This paper compares the marking of pragmatic roles in Central Somali oral narrative discourse and elicited question-answer pairs to illustrate the claim that information structure must be studied within the context of particular discourse genres. The study of pragmatic roles in Central Somali is especially interesting because clause-level focus is explicitly marked through the particle yaa. The functions of this particle in elicitation question-answer pairs and narrative discourse are compared, and it is shown that elicitation data exhibit only a few of the possible functions of yaa, viz. as a marker of assertive and contrastive focus. In contrast, narrative discourse provides examples of yaa as a marker of both event-clause focus and discourse topic. This result is discussed within the context of discourse coherence and is shown to be not as surprising as it first appears. In addition, narrative focus constructions (defined as the most salient section of new information in a narrative text) are shown to be formally well-defined and functionally important in giving coherence to a narrative, although no counterpart has been found in elicitation data. In conclusion, it is noted that pragmatic roles should be studied in a broad range of discourse genres in addition to elicited question-answer data, since each genre may illustrate different functions of the same constructions.

This paper presents a synchronic description of the realization of /l/ in Chumburung, a Kwa language of Ghana. In this description I attempt to demonstrate that the influence of the feature advanced tongue root is not restricted to the vowel harmony system, but also spreads to influence the realization of the consonant /l/. Following from this synchronic description is evidence suggesting that native speakers of Chumburung require more than phonetic information in order to choose the appropriate allophone of /l/.

Temne is a West Atlantic Mel language, spoken in northern Sierra Leone, which has both phonemic and morphological tone. This paper explores the interaction between tonal and segmental levels through the investigation of segmental rules of insertion and deletion and through verbal inflections and derivations. The paper shows that tonal patterns on Temne verbs are not additive, unlike the segmental portions of the verbs, and that the tones on verbs must be viewed as almost totally independent of the component morphemes of the verb. In the process, the paper argues for the analysis of an underlying vowel and for the analysis of pronouns as noncliticized morphemes, which means that Temne has redundant subject markers in certain types of sentences. Both of these are issues which various authors have taken opposing stances on in the literature.

1. *qamA "dry, warm" In my earlier paper I reconstructed the Nostratic word *qamA "dry, warm"; this etymological nest can be enlarged with some new examples: AA: Sem. *hmm "to be warm": Hebr. hām "warm", Ugar. hm, Syr. hammīmā, Akkad. emmu, Arab. hāmm; Eg. bmm "to be warm, to be hot" , Kopt. (Saĭd.) hmom "to become warm”;

The present paper is mainly concerned with the definite ar­ ticle of Beowulf. The literature about the West Germanic definite article has always allotted a prime place among the old Germanic texts to this poem: it is undisputably archaic (Kuhn 1933: 105, D. Hofmann 1959:189), and suitable for statistical genera­lisations. Its more than 6,000 hemistichs contain about 300 in­stances of the. definite article (Lichtenheld 1872:332).

Many African languages have a system of grammatical focusing which pragmatically highlights certain elements of a sentence. Such systems of focus often have significant consequences in the syntax and morphology of languages, in that selection of a particular type of focal morphology prevents syntactic rules from applying (or forces syntactic rules to apply). This paper investigates the focusing system of one Bantu language, Kimatuumbi. It is argued that the optimal account of Kimatuumbi focus is to allow the syntactic rules to apply blindly and to filter out the unacceptable conflicts in focus via a pragmatic filter.

This paper examines the different properties of what we have termed "auxiliary focus": the interaction between focus and the semantic features of tense. aspect, mood. and polarity. We argue that auxiliary focus has every property of its counterpart (termed "constituent focus") and that in order to account for focus of any type. we shall have to address the following four parameters: (a) realization of focus (prosodic, morphological. syntactic). (b) type of focus (assertive vs. contrastive). (c) scope of focus (subject. object, verb, auxiliary, etc.), and (d) control of focus (pragmatic or grammatical). A particularly interesting side of auxiliary focus is found under point Cd): while in some cases. in some languages, speakers are free to choose [+focus] or [-focus] auxiliarv markers according to the context (pragmatic control), in other cases the choice is dictated by the language itself (grammatical control). We will show that some semantic features of the auxiliary are inherently focused on universal grounds, providing a typology of focus marking on the auxiliary. The relevance of the above considerations of auxiliary focus to the development of tense/aspect systems provides the conclusion to this work.

Hausa ethnonyms, terms indicating a person's origin, ethnic affiliation, or professional or social position, are formed with a prefix ba-in the singujlar and a suffix -aawaa in the plural. This paper provides a detailed specification of the segmental and tonal characteristics of "these ethnonyms and a comprehensive list of currently acceptable forms. The paper illustrates the semi-productive nature of the construction and the imperfect pairing of forms with ba--/-aawaa. The use of ethnonyms to indicate 'supporters or followers of a person' is documented as well as the overlooked functioning of ethnonyms as adjectival qualifiers. The paper explores the relationship between the suffix and the identical suffix found with names of towns and villages. The major conclusion is that ba- and -aawaa are suppletive derivational markers, the former originally denoting 'person ... ', the latter indicating 'community ... '.

This paper takes its data from OshiKwanyama, a Bantu language spoken in Namibia and Angola; the paper is an investigation of the phonological incorporation processes at work in the loanwords of this language. When loanwords are taken into a language the degree to which they undergo such phonological modification varies. There appears to be a hierarchy of such modifications in OshiKwanyama, such that some always apply, others normally apply, and yet others apply rarely and in special circumstances. These modifications are often based on, though not necessarily identical to, the MSC's of the borrowing language. The modifications appear to work in terms of a hierarchy of application. The paper describes several loanword incorporation processes and delineates the modification hierarchy.

A current debate going on among Yoruba linguists is the existence and phonological status of the high non-expanded vowels. Indeed while Igala, Isekiri and many Yoruba dialects exhibit a seven-vowel system, other Yoruba dialects exhibit a ·nine-vowel system (including i and o), both types showing however interesting vowel co-occurrence restrictions. Given this situation, some scholars argue that proto-Yoruba and proto-Yorubold had a nine-vowel system with cross-height vowel harmony, while others claim that they had a seven-vowel system. The present paper reviews both positions and suggests another alternative; in particular it shows how cross-height vowel harmony came about in Yoruboid.

The replacement of the familiar linear phonological representations of the SPE-type by the multi-tiered configurations proposed in autosegmental phonology has enriched the theoretical apparatus available to the linguist and has opened the way for a better understanding of the workings of phonology. One particular strength of this theory is that it predicts the existence of distinct subsystems within a phonological system and offers a principled account of their interactions. Applied to the aspectual inflection of Silt'i verbs, the multi-tiered approach proves its potential by accommodating two distinct subsystems of inflection, viz. CV-pattern inflection and ablaut inflection, in a unified, principled description. From a more , general point of view, this paper sketches one possible way of integrating syllable theory as a distinct subsystem into an autosegmental system of phonology.

After a brief review of the literature on factivity and presupposition, the paper examines a case of asymmetry in the interpretation of a construction known as the relativised predicate (RP). Whenever the main verb of the sentence with RP is mek (= 'make'), the sentence may be interpreted in two ways, including one with "the fact that". However, if the matrix verb is other than mek, there is only one interpretation possible-without "the fact that". It is concluded that the dual interpretation is possible because the relative particle (which is also a complementizer linked to factive verbs) associates with the factive mek.

This paper examines the structural behaviour of various types of complex verbs in Nupe and Yoruba in relation to causative constructions. When such verbs occur in simplex non-agentive causative sentences as well as in agentive non-causative sentences, they freely permit sentence embedding, resulting in biclausal causative structures. But in the case of non-agentive causatives, it is only Yoruba which allows the verbs to be irregularly embedded into causative matrix sentences in such a way that the biclausal causative structure constitutes input to Causative Clause Union, which compresses it into a uniclausal agentive causative sentence. Functionally, Yoruba uniclausal and biclausal agentive causatives are intercharigeable in many cases. On the other hand, only uniclausal agentive causatives are attested in Nupe, as Causative Clause Union has disappeared from its grammar. It is concluded that causative constructions in these languages demonstrate clearly that the CCU rule is motivated by a diachronic process of moving from a pragmatic mode of expression to a syntacticized one, and where a particular target has been hit, the rule ceases to function.

Swahili demonstratives h- and -Ie have traditionally been analysed to mean "proximity" and "non-proximity" respectively. However, this analysis fails in that it can only account for a small part of the distribution of these forms in actual texts. This paper suggests that meanings dealing with the speaker's relative concentration of attention on a referent are better able to account for the actual distribution of these forms. To validate this claim we will (1) show the relation of proximity to noteworthiness and thus explain the same range of data as the proximity hypothesis; (2) uncover other factors, e.g. new items, thematically important items, that override proximity and show their relation to noteworthiness to explain data not accounted for by the proximity hypothesis.

Leaving aside Sancho de Elso's catechism (1561), the text of which has not survived the oldest Basque catechism is the one by Betolatza, Doctrina Christiana en Romance y Bascuence, published in Bilbao in 1596. We had no data about the author of this booklet. The purpose of this paper is to show that he was born at Betolatza (province of Alava) and that he was priest in his birth-place and in the cathedral of Santo Domingo de la Calzada, the town where he died between autumn 1599 and spring 1600. A copy of Betolatza's catechism has recently been bought by the Libray of Basque Parliement. Undoubtedly, this copy, the only one whose existence we know, is the same as the one Azkue photographed in the thirties.

This paper highlights the various studies which have been carried out to determine the rules which govern the elision of one of the vowels in a verb-noun concatenation where the verb is vowel-final and the noun vowel-initial. It proposes that there are basically three phonological rules which come into play at the Deep Structure Level. However, at the Surface Structure Level, the paper recognizes that there is a Semantic Dissimilation Principle (SDP), which may block the application of otherwise well-motivated rules. The SDP, then, guarantees maximal perceptual distance between otherwise homophonous products of the phonological rules.

This paper examines the need to look at quantifiers from semantic perspectives unconnected with logical, existential, or universal claims. The focus here is on the quantifier 'many' in Yoruba. Among other observations it is noted that the Yoruba quantifier meaning 'many' may occur only with those NPs that are viewed as animate.

In 1977, Bennett and Sterk published a reclassification of the Niger-Congo languages which has been highly influential. In this paper I try to discover their lexicostatistic method (section 1), then use their published data to do a conventional lexicostatistic subgrouping (section 2), and finally look at their evidence for denying the genetic unity of Narrow Bantu (section 3).

In recent surveys of tonal phenomena it is suggested that vowel quality rarely affects tone. In the present paper it is argued that one such case can be found in Lugbara. Internal and comparative evidence is presented that the vowel feature Advanced Tongue Root (ATR) has had an initiating role in tone splitting in Western Lugbara. It is demonstrated tha.t the pitch of a vowel with a high tone was raised if the vowel was [+ATR], and it is suggested that the resulting tonal allophony was later phonemicized due to vowel merger and reduction or loss of segments.

This paper examines the effect of [ATR] vowel harmony on low vowels in Okpe, an Edoid language of Nigeria. The relevant facts can be summarized as follows: Low vowel stems condition [-ATR] forms on affixes. Low vowel affixes surface as [+low] in [-ATR] contexts and as [-low] when in [+ATR] contexts. Of particular interest is the additional fact that an underlyingly low vowel surfaces as [-low], [-ATR] in certain [+ATR] environments. To explain these alternations, it will be argued that low vowels are underlyingly unspecified for vocalic features. Redundancy rules, supplied for the most part by Universal Grammar, interact with the vowel harmony system and rules of syllabification to derive the non-low variants of under1yingly low vowels. By positing underspecified forms, it will be shown that no ad hoc rules need to be stipulated.

L'Università Autonoma di Barcellona ha inserito fra le sue Publicacions del Departament de Filologia Hispànica il volume che reca il titolo citato e che è desti­ nato ad avere carattere periodico (p. 6). Nel Prologo (pp. 5-6) i curatori affermano che fino agli anni 70 la linguistica «en aquest nostre país» non era arrivata al grado di maturità che permettesse la formazione di una scuola né di un sufficiente numero di ricercatori. Dall'inizio della scorsa decade le cose sono cominciate ad andare me­ glio: nel giugno del 1983 è stato spedito un invito a collaborare al volume miscella­ neo (qui recensito) e a Sitges è stato o ganizzato un colloquio internazionale dedica­ to ai problemi di linguistica teorica e agli aspetti (sincronici e diacronici) della lingui­ stica romanza che presentano interesse teorico. Il volume recensito raccoglie tredici contributi, di cui sei sono stati presentati al colloquio citato. Con le parole dei cura­ tori (collaboratori dell'Università Autonoma di Barcelona /U. A. B./: Joan Masca­ ró, Anna Bartra, Josep M. Brucart, Josep M. Nadal e Gemma Rigau): «d'aquesta manera volem fer plataforma que contribueixi a la consolidació, al nostre país, d'aquella pràctica científica "normal" que esmentàvem abans» (p. 6). Insomma, si cerca, e si riesce, a portare la linguistica catalana ed ispanica ad un livello internazionale e nel contempo si presentano i suoi risultati al pubblico linguistico mondiale.

Raising in Berber has been selected for study for two main reasons: firstly, this syntactic operation which has hardly received any attention up to now involves areas which are central to the understanding of the whole syntactic structure of Berber. Secondly, Raising in general is considered by Chomsky and most linguists as being a universal feature of natural languages. However, although Chomsky does not question the existence of Raising to Subject, which he subsumes under the general rule of NP movement, he seems to strongly reject the rule of Raising to Object, claiming that what this rule does can easily be carried out by means of one of his constraints, the "Tensed S Constraint". The goals of the present paper are twofold: the one hand, it aims to show that although Chomsky considers Raising to Subject as an essentially movement rule, this rule seems to operate in Berber without syntactic movement. On the other hand, the paper aims to show that Chomsky's theoretical suggestions in the domain of Raising to Object are highly questionable, if not utterly wrong. A number of arguments supporting these two views are based on Berber data.

This paper examines a category of compound verbs which features a particular kind of affix morpheme, with a view to determining the nature of the relation between this affix and other homophonous forms in the language which synchronically belong to different grammatical categories. In line with findings from similar studies of serial verb constructions in languages such as Mandarin Chinese and Yoruba, we have shown through our Edo data not only that certain verbs undergo grammaticalization and lexical re-analysis historically in the context of serial constructions to become prepositional case markers and adverbs synchronically, but also that some of these grammaticalized forms do undergo further lexical reanalysis and semantic depletion and ultimately become incorporated as affixal morphemes in compound verbs.

The goal of this paper is to show the present growing and flourishing of Acoustic Phonetics thanks to the perfectioning of electroacoustics and its tools (Sonograph,Synthetizer, etc.). It starts with some general observations about communication, the car and the sound, as well as the type of material that the modern use of electrotecnic methods makes available. The main part of the paper describes and characterizes some vocalic and consonantic phonemes of both Basque and Castilian Spanish acoustically by means of many data and illustrations.

This paper examines the strategies for relativisation and the Noun Phrase accessibility hierarchy and constraints in Yoruba [Keenan and Comrie 197T]. The various positions relativisable are examined. It was found that contrary to what Keenan and Comrie thought, Yoruba relativises all positions except the Object of Comparative. Attention is also focussed on the status of the coreferential pronoun found in subject relativisation. From the presence of this pronoun it appears as if Yoruba violates the continuous segment principle. However, our analysis shows that the pronoun is a surface structure phenomenon which obscures the underlying strategy for relativisation. The Yoruba data therefore supports the Hierarchy Constraints.

This paper looks at the strategies employed by speakers in Topic Shading in an unplanned Igbo discourse. The results of an investigation involving some Igbo adult students might lead us to infer that individuals shade topics for various reasons and that conversational rules may not be as obligatorily governed as those of grammar. Shading strategies might depend on culture and context. A cross-linguistic study of planned and unplanned discourses might help not only in developing more appropriate and specific conversational maxims, but will also be helpful in determining how strategies of conversational interaction could be integrated into other aspects of speakers' linguistic knowledge and culture.

The paper considers four alternative treatments of the grammatical status of the adjectival forms gera, gražu which occur in Lithuanian grammars: as the neuter forms of the adjective, adverbs, statives, and the non-gender forms of the adjective. It is argued that the existence and use of these forms, just as the existence and use of the masculine and feminine forms, can be explained adequately by laws of syntactic agreement. Masculine and feminine adjectival forms are used when the grammatical subject is a nominal with some gender meaning, i. e. a noun or a pronoun, forms with the ending -a, -u are used predicatively in subjectless clauses or when the position of the grammatical subject is filled by a nominal which does not possess any meaning of gender, i. e. an infinitive, a non-gender pronoun, a subordinate clause or some other phrase. It is logical, therefore, to treat the forms gera, gražu as the non-gender forms of the Lithuanian adjective.

The purpose of this paper is to review several terms spread over a quite large area in South-East Europe. The starting point of our investigation is the Romanian language understood as inheriting an important Thracian vocabulary, specifically referring to the social and political structure of the Early Middle Ages. The terms discussed are not exlusively Romanian. In fact, they reflect - roughly speaking - the ancient extension of the Thracian speakers, i.e. the present-day territories of Romania, Bulgaria, Soviet Moldavia and parts of South- and South-West ·ukraine, Slovakia, Hungary and Yugoslavia.

This paper presents evidence that at least three languages in the Surma group (and possibly also Proto-Surma) are exceptions to the proposed typological "universal" that languages with basic VSO word order do not have post-positions [Greenberg 1963:78]. There is no attempt made here to hypothesize how this came about, only to call attention to the fact that these three languages are typologically anomalous.

Contrary to the view that in Bantu languages the two unmarked nominals following the verb in ditransitive constructions need not be distinguished because both possess the same object properties, this paper shows the necessity of making a distinction between the direct object and the indirect object relations. Evidence comes from SiSwati, the language of Swaziland, and the analysis of the data is cast in the Relational Grammar framework. The arguments presented refer to word order, object concord (or pronominal copy) and the interaction between object concord and some syntactic phenomena such as passivization, topicalization, relativization, and clefting. By distinguishing the direct object from the indirect object in Siswati, the grammar is able to provide a more natural account for a number of related double object constructions.

In this paper some propositions are made concerning the analysis of Basque relative clauses. In the first part, it is shown that the relativized position is better analyzed as being filled by an empty operator, than by an empty pronoun. The arguments are as follows: i) the fact that there is only a woolly correspondance between the licensing of pros and accessibility to relativization; and ii) the evidence of the cyclical movement of the relativized NP in complex relative clauses. In the second part, relative clauses containing a WH-word are examined, ie. questioned relative clauses and relativized embedded questions. Assuming that the empty operator in relative clauses is moved to COMP to the same A'position as WH-words move to, it will be shown that both constructions can be explained straightforwardly, without violation of Subjacency, allowing however some adaptation of the Double-Filled Comp Filter.