Band 9, Heft 1

(a) Within the methodological approach of Natural Morphology we evaluate the model of the 3rd declension built up on implicational relation among morphological properties as proposed for masculines and feminines of the 3 rd declension by Wurzel (1989). We comment on Dressler's (2001) potential system of Latin inflection and the classification of default and non default forms within the macroclass of the 3 rd declension nouns given by him. We add the model of the macroclass of the 3 rd declension adjectives. Some criteria of productivity of the consonantal, i- and mixed classes are dealt with, as well, (b) Referring to material in Grammatici Latini (Priscianus) we observe the functioning of extramorphological (phonological, semantic and morphosemantic) properties. Within Priscianus' model one extramorphological property or a simple combination of properties does not predict paradigm structure conditions of only one and not some other class. Nor can we rely on Priscianus' implications among morphological properties, (c) In the 3 rt declension nonfunctional formal distinctions tends to be eliminated. The dominant paradigm structure conditions: acc. sg. -em, abl. sg. -e/-i, gen. pl. -um/-ium, acc. pl. -es are formed. The fact that Latin became a dead language might have been taken into consideration in simplified inflectional rules used at school: only the dominant paradigm structure conditions of 3 dominant classes for masculines and feminines and 2 classes for neuters are related to extramorphological properties; the network of "exceptions" has to be added.

The Latin suffix sê (imperfect of subjunctive) has no parallel apart from Italic languages. Most of the explanations given by comparatists are based on the hypothesis that such suffix is bom from the coalescence of s (future or aorist) and ê, which has something to do either with *yeH (optative) or with *ê (long vowel subjunctives). Relying on the fact that Indie ya answers to Latin ê in denominatives and statives, the present paper tries to establish a link between Latin subjunctives in sê and Vedic conditionals in sya. Those conditionals signify what was going to happen, what should happen if ... ( qbharisyat "was going to bear off'), which is not so far from the meaning of Latin imperfect subjunctives.

In recent theoretical literature on morphology, the question has been raised whether there are languages in which some morphological device exists, which turns out to be the "subtractive" counterpart of the comparison of majority, according to the following proportion: great : great-er = more great :: great-X = less great, where X corresponds to a morpheme meaning 'less'. In this paper a tentative analysis is put forth to show that, if any, the likeliest candidate to exercise this function is the Latin preposition sub- in composition with adjectives.

Compared to the 'regular' subjunctives, the sigmatic ones (types amassim, habessim and faxim) do not have a large number of tokens in Early Latin. After that, they survive only as deliberate archaisms. We can see their decline already in the pre-classical period: Plautus has many tokens belonging to a variety of verbs, while the formation is clearly fossilized in Terence. As early as Plautus, these subjunctives are not used much outside main clauses, and forms o f facere and audere prevail. The rarity of the sigmatic subjunctives goes hand in hand with higher register. As far as the temporal characteristics of the sigmatic subjunctives are concerned, they behave like present tense forms and the non-past perfect subjunctives employed in prohibitions and potential main clauses. The sigmatic forms have remarkable distributional properties. They are disproportionately frequent in prohibitions and subordinate ne-clauses, but absent in commands or exhortations addressed to the first person plural. Besides, prohibitions with sigmatic forms are almost exclusively used in the second person. The sigmatic subjunctives share these restrictions in main clauses with the perfect, but not the present subjunctives. This points to an aoristic rather than a desiderative origin of our forms.

In this paper I analysed the language of the Latin official inscriptions discovered in the central-eastern provinces of the Empire (from Raetia to Egypt). I restricted my study to those texts which contain the name and the titles of the emperor ( tituli honorarii, tituli operum locorumque publicorum and miliaria). This analysis revealed that over 80% of the morpho-syntactical deviations occurring on the epigraphs come from Greek or bilingual areas. Specifically, most of them can be traced back to three reasons: 1) The necessity of sparing space on the stone; 2) The confusion between different formulas; 3) The influence of the Greek formulary.

The fragments of Caesar's De analogia that deal with nominal morphology can be viewed as proposals for a rearrangement of inflectional paradigms. This is especially the case with the paradigms of /-stems, the independence of which was compromised by the (partial) fusion with consonant stems, and those of u- and e-stems, which were little attested and little productive. Caesar tries to level out differences among morphemes and to promote syncretism of inflectional forms, because he intends to single out the individuality of each nominal class. A comparison between Caesar's theoretical statements and literary evidence shows that he strives to resist language's decadence insofar as he abides by the past consuetudo pura et incorrupta. In general, Caesar's proposals aim to observe good usage: among the possible options offered by the literary tradition only those which can be best inserted into the linguistic system are chosen.

After questioning the examples of inscriptions in the Roman period which have been adduced in favour of Id and loi diphthongisation, the author deals with the situation of these vowels in Hispanic territory in two different periods: first in Visigoth texts, in which no evidence is found in support of the abovementioned diphthongisation, and secondly in diplomatic documents from north-western Hispanic territory, in particular the kingdoms of Asturias and León (718 to 1037). The latter shows the first irrefutable evidence. The scenario described proves that diphthongisation is not a spread phenomenon and it seems to be less frequent in the case of loi than in the case of lei; there also seem to be differences related to the documents studied and geographical area. As for results, 'ie', 'ia' for etymological Id and 'oai' for etymological loi are attested only in free syllable.

The aim of this paper is to draw a sketch of the verbal voice system in Latin and possibly to shed more light on some controversial points (in particular, the status of deponent verbs (DVs)), by means of a comparison between middle voice (MV) and Split Intransitivity.

After reading the palliata comedy, we have checked that the frequentative suffix is very often used by Plautus and not so much by Terence. Starting from that consideration, we have analyzed the -itsuffix, reaching the following conclusions: (1) This suffix appears in the Latin theatre mainly in genres genuinely Roman such as atellana or togata. The overwhelming majority of the verbs concerned with the corpus analyzed are not or very rarely documented in the latin language, except for certain writers. This confirms the fact that -it- comes from an italic suffix which was extensively introduced in the Latin language on the 3rd century b.C. (2) The generalized use of these verbs in the theatre shows us that they were regularly used in ordinary language, being the reason why there is hapax which comes directly from the Greek language (e.g. noscito ) and from the present (e.g. crepitai) in a regular way following the first conjugation paradigm. (3) The "short-long" sequence in agreement with the suffix "it-α" can often be found in final position in dramatic verbs or coinciding with caesura. (4) From the examples analysed the hypothesis that the origin comes from the past participle and not the supine seems to be confirmed.

Three parameters are considered in the typology of Latin syllable structure: 1) phonotaxis of syllable structure; 2) correspondence of syllable structure with morpheme structure; 3) syllable structure processes. With respect to these parameters, Latin may be characterized as follows: 1) Latin allows moderate complexity of the onset, which, during its history, gets further reduced. Similarly, complexity of the nucleus is mostly reduced to a single short vowel. Latin allows relatively complex rhymes, which, however, get reduced in the language history; 2) Latin (as well as Romance languages) goes relatively far in blurring morphological boundaries, even left stem boundaries and, thus, word boundaries, by 'liaison'; 3. Latin phonology is relatively faithful to lexical representations. However, syllable structure is simplified, at the expense of the phonological manifestation of grammatical boundaries.

In this paper, the history of the studies on Latin nominal compounding is outlined. In the first part, Th. L. gives an overview of the development from the ancient grammarians to the beginning of scientific research in 19 th century. In delineating the progress achieved by the mostly philological approaches, also the comparative and linguistic contributions are included in the discussion. It was particularly through the neogrammarian perspective that some aspects of Indo-European word formation became also an important factor for studies on Latin compounds. In the second part, R. O. discusses the main innovations that took place in 2 th century. While the historical research in the traditional fields of typology and stylistics went on, two new paradigms of general lingusitics shed new light on the synchronic side of compounding. The first attempts to investigate the phenomenon of Latin composition as a system were elaborated by structural linguistics. This perspective was carried on by generative grammar, whose most important progress was the formulation of word formation rules and the discussion of their interaction with other components of grammar.

After more than twenty-five years from the discovery of the Lapis Satricanus - in 1977 in the foundations of the Mater Matuta temple II in Satricum - many matters related to the epigraph are still open. One of the most important issues concerns the placement of the inscription inside the linguistic diasystem of Central Italy. In this paper, a short version of Lucchesi-Magni 2002,1 focus my attention on this aspect specifically, evaluating all previous hypotheses and then trying to present an alternative solution based not only on the historical and cultural data of the Lapis Satricanus, but also on the archaeological, epigraphical and, of course, linguistic ones.

In the first section of this article I see that in -fico(r) verbs the presence of deponent is connected with the substantival nature of the first element and with the non-factitive meaning of the verb. In the second, I show that the distinction between active and deponent reproduces the distinction between the light verb facere [+ causative] and the light verb facere[- causative]. In the third section, I consider the properties and the role of the subject of -fico(r) verbs and the subject of light verb facere constructions, so that I'm able to find out as common features [+human] and [+inactive] traits. Lastly, in Plautus the opposition between ludifico and ludificor can be explained with a scalar evaluation of the [+ inactive] trait.

The aim of this study is to refute the standard etymology of the placename Latium on the one hand and to propose a new etymology on the other. The standard etymology is semantically 160 years old and formally more than 90 years old and traces Latium as *(s)tls-t-/o-m 'flatland' back to an ie. etymon *stelHr (trad. * stela- : 'ausbreiten, flach hinbreiten') or *telhr- (trad. *tels- : 'flach, flacher Boden, Brett'). Linguistical and philological considerations, however, preclude the possibility of this etymology and make room for a new one, i.e. for the deduction of the placename Latium as *latjom 'marshland, moorland' from the ie. etymon *lh 2 t- (trad. *lat- : 'nass, feucht').

In my paper I describe the vocabulary used in the Paris manuscript of the Apocalypse of Paul, and I scrutinize especially four words: quia, quoniam, super and usque ad. The notable features of the grammar (phonology, morphology, syntax) of this manuscript are already discussed, but their vocabulary is neglected in the special literature. Inspite of this fact, the study of the vocabulary of the Apocalypse of Paul is very instructive, concerning the origin of the manuscript. I carry out my research on the basis of the edition of Theodore Silverstein and Anthony Hilhorst: Apocalypse of Paul. A new critical edition of three long Latin versions. Patrick Cramer Éditeur, Genève, 1997.

Proper names are specific to the civilisation and to the language in which they occur. The 'translation' of a proper name involves in the first place establishing the identity of its referent This identification is conditioned by the degree of familiarity with the proper name in question and by the cultural and linguistic level of the speakers and hearers who use it. It is often based on a presentation strategy which has recourse to characterising expansion and metalinguistic paraphrase. The translation of a proper name also involves a confrontation with the structures of the target language into which it can be assimilated to differing degrees. In most cases proper nouns become loanwords, but it can also happen that they are translated, that is to say replaced by lexical equivalents, of a greater or lesser degree of complexity, which already exist in the language or which are invented, often maladroitly, for that specific purpose. These translations display a concern to explain rather than to name the foreign object. They reveal differences in status between different onomastic categories and lead to a consideration of the controversial question of the 'meaning' of proper names*.

The aim of this article is to state that the constituents of three Latin phrasal lexemes: custos corporis, (re)linquit/-itur + animus/animo and de celerò, experience a progressive integration in writers preceding Quintas Curtius Rufus, until they appear as fixed lexicalized forms in this author, and afterwards each of them has a different destiny.

Latin gradable adjectives can be modified by prefixes and suffixes which can be intensifies or disintensi fiers. They have the semantic particularity of showing the notion of a scalar progression, a subcategory of quantification. They could be called scalar morphemes. The intensification may be either towards a high or a low degree of realisation of the notion. We study here three productive scalar morphemes : -cuius /-ulus, sub-, per-.

Privative oppositions have arisen with Phonology within the Prague School and they are a very fruitful structural concept. Their operative power has been successively verified on the morphematic, syntactic and lexical levels. Whereby, they are also a good proof that the systematic character of the language is not limited to the grammar. A privative opposition always contains three values, namely positive / negative // neutral. These can be expressed by three lexemes (Lat mulier / uir II homo; Germ. Frau / Mann II Mensch) or by two lexemes (Lat nox / dies II dies', Sp. mujer / hombre II hombre ; Fr. femme / homme II homme-. It donna / uomo II uomo)·, in this last case, one of them indicates the negative and neutral values. Even though a privative opposition is a complex structure, it is more precise than the concepts of synonymy ( homo / uir), antonymy ( uir / mulier, dies / nox) or polysemy ( dies H dies). The values of a privative opposition clearly reflect the three phases (thesis, antithesis, synthesis) of the Hegel's dialectic process, but their historical development is often inverse.

The aim of this paper is to consider what is a "problem" in Latin from a cognitive perspective: a problem is not but "a thing that falls on us", "a place without exit", or "a thing without movement". We shall study etymologies of Latin words for "problem" and "solution" and usual expressions such as haeret haec res, or vadum safaris, where it is possible to find Source Domains (places, exits...) by which Abstract Concepts can be established as Target Domains to express the complex concept of problem, according to spatial and ontological metaphors.

Medical and veterinary texts use the verb facere with a meaning which is often peculiar to them. With a name of an ingredient or a disease, facere forms a periphrasis which compensates the lack of a simple verb (facere pastillos, facere equum hydropicum), or avoids the use of a too vulgar verb (urinant facere = meiare). The present participle or the third person singular (faciens ad, facit ad) means "efficient" for a disease. Finally, facere introduces an infinitive clause while having a factitive meaning, in which it is equivalent to a jussive-verb. This construction, still rare in the 4lh century, is the ancestor of the French expression "faire faire".

The problem dealt with in this paper is the extent to which the changes in tense, viewpoint and actionality in Late Latin are reflected in different kinds of texts. The changes in the actional system lead to the introduction of new words, of which some are accepted in literary Late Latin and others are not, and to the use of old words in partly new functions in the literary texts. The use of fueram instead of eram with past participles in forms indicating the pluperfect and the employment of the imperfect rather than of the perfect in expressions which give an overview of a past situation, e.g. with durative temporal adverbials, is found in most kinds of Late Latin, but is considerably less frequent in the literary texts. Not even the most learned Late Latin authors understood - and avoided - the changes in the actional system, but to some degree they seem to have been aware of the changing functions of the tense forms.

This paper takes as a starting point the idiomatical use of quasi in lexical explanations, e.g. hodie quasi hoc die, scabies quasi scamies, etc. which occurs frequently in philological and lexical literature, such as Servius and St. Isidore of Seville, whose work together with that of Macrobius forms the basis of this investigation. It is argued that quasi in such instances should be taken as "in the sense of1 rather than "as i f ' and that it probably developed from the use of quasi in the sense "in the capacity o f ' and related causal uses of quasi (without a finite verb), which may descend from the finite uses of this conjunction. The first instances of the lexical use of quasi in Varrò, Apuleius and Gellius are discussed. In Macrobius, Servius and St. Isidore, quasi developed further functions and occurs far more frequently than in earlier authors of the first two centuries A.D., but it is less connected with finite clauses. Parallel developments with tamquam, which was used more frequently by early authors, are briefly discussed.

This paper deals with the problem of defining and describing semantic roles. Methodologically, I follow Coseriu (1987), who argues for a sharp distinction between language-independent reference roles and language-specific semantic roles and criticizes earlier approaches which do not make this distinction. Recent contributions to the problem of defining and describing semantic roles or functions within Generative Grammar (Jackendoff 1991) and Functional Grammar (Dik 1997) differ from earlier approaches in that they try to take this distinction into account However, some problems remain. When we have to deal with syntactic and semantic contrasts between languages such as Latin and German, these contrasts cannot adequately be described with the help of language-independent theta-roles or semantic functions. Therefore, it seems preferable to adopt Coseriu's strategy of describing the expression of reference roles like 'Possessor' in Latin and German as a set of contrasting contextual meanings of cases, syntactic functions and constructions. Contrastive data show that the relevant constructions in Latin and German differ widely both as far as their number and their functional properties are concerned.

The hapax legomena substantives and adjectives created with a suffixal derivation by Lucretius are generally shaped according to the established derivational rules. From a cognitive point of view, these words corroborate that the two main activities of the human brain worked out in the natural languages are selection and association. Some less expected words allow us to point out an important parameter of the lexical creation : the pertinent features are set in a hierarchical frame. The remaining question is to know whether this hierarchy is inscribed in the system of each language or depends, entirely or partly, on the free speaker's choice.

In the recent literature on grammaticalization we find that, since modal meanings arise from a restricted set of semantic sources (i.e. lexical items that refer explicitly to concepts related to possibility, necessity, etc.), cross-linguistically similar paths for the evolution of grammatical markers are predictable, and multiple uses of forms can be viewed as points on grammaticalization chains. The purpose of this investigation, which takes into account the data provided by extensive typological research and the theoretical framework offered by Role and Reference Grammar, is to test these findings and predictions about universal modal paths. Starting from the analysis of the modal forms possum, licet, debeo, oportet , and necesse est , I will also propose a semantic map of modality for Latin.

Within the field of structural semantics, there has been a series of excellent contributions to the subject of lexical gaps (places in the lexical structure of a language where we would expect to find unity-but do not find it). The theoretical work of H. Geckeler and others, however, has not been applied systematically to the study of Latin. In this article we present a first approximation to this subject.

This paper proposes a new analysis of the Latin perfect studied from its use in the St. Augustin's Confessions : it is an attempt to clarify the aspectual value of this tense in an enunciative perspective (Culioli's theoretical framework). It is well known that the Latin perfect occurs either as an accomplished tense or as a revolved past tense. Both these contextual values may proceed from the only following abstract meaning: the Latin perfect has to be used when the speaker wants to show that the terminal boundary of the process is overleaped and gone beyond to its complementary result The status and the interpretation of this resulting situation vary with the contexts; so we analyse the different contextual parameters that constrain the interpretation. The main issue under discussion is to decide whether the operation of temporal location built by the Latin perfect needs necessarily a locator (reference point), especially in the enunciative situation, or whether it itself is its proper origin. We'll assume that the Latin perfect is not a deictic tense, but a very aoristic one.

The aim of this paper is to investigate a Latin lexical field (the field for forgiveness) which has progressively increased. In addition to ignosco and uenia, many polysémie verbs and nouns have enlarged their semantic area with the meaning "to forgive" or "forgiveness". These words, which are still polysémie, are partly synonymous.

Varrò in De lingua latina was the first to draw a distinction between authentic Latin words (verba nostra) and lexical elements of foreign origin (verba peregrina). The aim of this paper is to examine the different attitudes of Roman authors and grammarians towards foreign elements in general. Specific attention is paid to the various arguments against lexical borrowing (lack of comprehensibility, offence to Latinitas, corruption of style by barbarisms, etc.) as well as to the role of grecisms in Latin linguistic thought Among the factors that have brought about a gradual change from interdiction to integration, five are discussed more closely: (1) the poetic licence (sermo poetarum); (2) the growing need for technical and special terms; (3) the general authorization to use words from every part of Italy (Quintilian); (4) the increasing association of latinitas with consuetude ; (3) the revolution of linguistic theory caused by Christian authors. Finally, the paper shows how grammarians in the late Latin period turned away from banishing foreign words and preferred only to set up rules of fitting them into the traditional language system.

The main aim of this paper is a linguistic study of agrimensor, a latin word among many others (ffnitor, decempedätor, grömaticus, mëtâtor, mênsor, togâtus Auguslôrum, etc.) for a land-surveyor. It is pointed out that agrimensor is a compound noun, a morphological and semantic copy of the greek γεωμέτρης.

In this paper I focus on the contexts that might have favoured the reanalysis process which leads the sequence habeo + object + a predicative passive perfect participle to periphrastic constructions with auxiliary habeo. Such contexts can be observed already in archaic Latin, in instances where the function of habeo is not connected with the expression of possession. The background that allows the grammaticalization of habeo and the development of periphrastic forms is, rather, the semantic decoloring of the verb, which seems to be well under way at Plautus' time. Traditionally, the first instances where the reanalysis took place are seen in occurrences with passive perfect participles belonging to the class of verba sentiendi (with examples from classical prose), but I use Plautine examples to argue that a relevant role has been played by specific passive perfect participles whose lexical content is such that the meaning of the whole sentence is in contrast with other semantic relations between habeo and the object.

This study presents a new proposal for the so far uncertain etymology of Lat. caballus. During the investigation based on the critical analysis of the earlier hypotheses (substratum, Wanderwort ) the author clarifies its basic meaning and its exact relationship (Old Greek χαβάλλης ; Lat. cabo; Old Church Slavonic kobyla; Proto-Germanic *χερα- [as a loan in Balto-Finnic languages]), especially its Asiatic cognates (New Persian kawat) and adds a new one (Khotan Saka kaba-). These are regular reflexes of a PIE stem *keb - *kob- "horse" suffixed with *-o- and *-ön. Caballus itself is a loan from an Old Eastern Iranian dialect, which has been passed on to Italy by the horse trade between the Carpathian Basin, Greeceland and the Venets. The word caballus marked the object of this trade, a new kind of horse introduced by Iranian tribes from the Steppe. This phenomenon is very well supported by the archaeological evidence.

An analysis of the introductions and conclusions of all instances of direct speech in Virgil's Aeneid (book I - VI) reveals that in these introductions and conclusions the present tense is used to describe events that are contemporaneous with "the effective span of story time", i.e. the moment in the story world that is narrated at that point in the narrative. The effect of this is that the narrator takes up a position inside the story world, from which he describes what he sees as the story time proceeds. This type of presentation is called the immediate mode. The use of the perfect tense is a more complex matter. On the one hand, the perfect is used as the main tense of (a particular part of) the narrative: the narrator presents his material from his own position in time. In this paper this retrospective presentation is called the detached mode. On the other hand, isolated perfect tense forms also occur in the immediate mode. In such cases the perfect is not just anterior to the speaker's time (as is the case in the detached mode), but has to be interpreted as anterior to the effective span of story time.

This paper concerns certain uses of the Latin accusative where it does not behave as an object, but as a subject. Some of these Accusatives, especially the proleptic accusative and the exclamative accusative, seem to share some of their most defining characteristics with the construction known as 'Nominatiuus Pendens namely, being separated from the rest of the sentence by a pause. Therefore, we think that the existence of the 'Accusatiuus Pendens' could also be recognised in Latin. The paper concludes with some reflections on subject accusatives of the infinitive.

This paper presents a possible approach to research on Latin clitics that is based mostly on statistic observations. The research was performed by observations on frequencies of different word lengths in various Latin texts. The results show a quantitative predominance of shorter words in Latin texts of more colloquial nature. They also demonstrate that shorter words in Latin have a predominantly grammatical function.

This paper starts from the consideration that, when an indefinite pronoun is found in a Latin conditional clause, it is quis in most cases, but in some cases it can also be aliquis or even quidam. Our investigation tries to explain both the usual case represented by quis and the exceptional cases, represented by aliquis. Furthermore, we will take into consideration negative conditional clauses, trying to explain unexpected phenomena like the low number of occurrences of quis and the absence of quisquam in nisi clauses.

The aim of this paper is to analyze whether, besides the basic functions traditionally ascribed to the three fundamental types of subordinate clauses in Latin, each of these subclasses cannot occasionally present functions characteristic of both other classes. Without any pretension to exhaustiveness, we will show by a sample of examples, gleaned from ancient, classical and Silver-Latin, that the subordinate clauses introduced by synchronically non-ambiguous forms of relative pronouns or circumstantial conjunctions can be functionally polyvalent, whereas clear-cut substantive clauses seem to be regularly confined to one specific functional domain. Various syntactic and semantic factors, considered both from a diachronic and synchronic perspective, prove operative to explain such differences of functional behavior.

The aim of this paper is to complete Maurel's (1989) analysis on the so-called Relative Verschränkung of the German grammars. I examine the differences between two types : a) relative clauses containing a bridge-verb introducing an Acl, an wz-clause or an indirect question b) relative clauses containing a second relative clause or another subordinate clause. I take into account the frequency of these two types, the constraints on the form and the position of the subordinate clause within the relative clause.

The paper surveys a characterization of diverse constructions of Ju/n+Dative from different linguistic levels (Semantic, Syntactic and Pragmatic). Two fundamental questions are particularly dealt with: the syntactic insertion in the hierarchical structure of the clause and the possible SF of the Datives examined. The proposed solutions take into account: (i) the graduality affecting the presence of different Arguments in nuclear predication and (ii) the procedure of prototypicity in the SF suggested, especially in the case of Recipient/Beneficiary.