Southern Salentino varieties have a five-vowel system, /i, epsilon, a, (sic), u/, exhibiting harmonic alternations in stressed mid-vowels: /epsilon/ and /(sic)/ raise to the mid-close [e] and [o], respectively, when followed by high vowels [i] and [u]. This harmony process is traditionally referred to as metaphony. Southern Salentino metaphony is characterized by variation in the way in which the unstressed [i] and [u] affect the stressed vowels /epsilon/ and /(sic)/. We examined the Southern Salentino dialect spoken in Tricase, focusing on both interspeaker and intra-speaker variation. Our primary goal is to understand the acoustic and articulatory characteristics of the metaphonic patterns and identify the features involved in this assimilatory process. Acoustically, we found that the metaphonic variant [e] has lower F1 and higher F2 values than /epsilon/, and the variant [o] exhibits lower F1 and F2 values than /(sic)/. We also show that the articulatory feature [+/- ATR] is significantly correlated with the metaphonic contrasts. Our second goal is to develop an appropriate phonological analysis of the phenomenon and describe the interacting parameters that generate the observed metaphonic variation. We argue that a key factor in the variation is the constraint against [+back, +ATR] vowels, which prevents high [+back, +ATR] triggers and/or mid [+back, +ATR] outcomes. After ruling out sociolinguistic factors, we propose that, although phonologized, the metaphonic process is still not fully normative in the community's speech, leading to both interspeaker and intra-speaker variation.
Homophony is often removed from languages to improve form-function isomorphism; this removal is known as homophone avoidance or ANTI-HOMOPHONY (AH). This article examines diachronic data from Aragonese and argues that AH, mediated by resource redistribution (ostensibly, suffix-borrowing), played a key role in the development of its verbal morphology. Of primary interest are the 1st singular (1s) and 3rd singular (3s) forms of the Pyrenean highland varieties of Ans & oacute;, Panticosa, Tella, and Benasque. While various authors have proposed that contrastive 1s and 3s have emerged as an AH effect (e.g., Ansotano tenebay similar to teneba; Benasqu & eacute;s teneba similar to tenebe; Tellano tenebe similar to teneba; as opposed to Panticuto teneba similar to teneba 'I had similar to he/she had)', AH as an organizing principle in Aragonese has not been rigorously explored, nor have the implications of AH for morphological structure been systematically evaluated in a more general sense. This study argues that AH in Aragonese draws upon distributional resources that vary across dialects, thereby mediating or impeding the resolution of 1s3s homophony. Verb data reveal how the absence of resources allows homophony to persist, though presumably in a state of tension.
This paper reconsiders Latin liquid dissimilation, which since Steriade, Donca. 1987. Redundant values. In Papers from the 23rd regional meeting of the Chicago Linguistics Society, 339-362. Chicago: Chicago Linguistics Society, has been understood as the change of the suffix -alis into -aris when there is an [l] in the base a process that is blocked by an intervening [r], but not by other consonants. Cser (2010/2020) observes that intervening noncoronal consonants also block dissimilation, for which Roberts, Philip. 2011. A proposed Latin rule insertion revisited in OT. Paper presented at the 85th annual meeting of the 85th Annual Meeting of the Linguistic Society of America, 6-9. Pittsburgh: Linguistic Society of America, proposes an OT account. We critically discuss Cser's empirical claims and Roberts's theoretical account and show how the blocking of Latin liquid dissimilation by dorsal consonants, but not by labial ones, can be made to follow straightforwardly from phonological structure.
In Spanish, the most consistent correlate of lexical stress appears to be duration. Even in phrasal contexts where the pitch contour is flat over all syllables in the word (as in parentheticals) and where pitch prominence is displaced to a different syllable (under rhetorical stress), lexically stressed syllables are usually durationally enhanced. Given this, the phenomenon known as "tonada cordobesa" is of particular interest, since it involves the lengthening of the pretonic syllable, which reportedly may become considerably longer than the tonic. The question that arises is what other acoustic cues are used to convey information on the lexical prosodic pattern of the word when increased duration is disassociated from the stressed syllable. To answer this question, we acoustically analyzed relative duration and pitch of 35 words produced by eight speakers native to different cities in the province of C & oacute;rdoba, which were compared to those same words as produced by six natives of the province of Buenos Aires, where pretonic lengthening is not expected. Results revealed that tonic syllables were significantly longer than non-tonic ones in both varieties. However, the tonic-pretonic difference was smaller in C & oacute;rdoba than in Buenos Aires, indicative of pretonic lengthening. Differences in pitch change between a pretonic and a tonic were not significantly different between provinces, but an overall larger pitch change between tonic and non-tonic syllables was obtained in C & oacute;rdoba. Together, these results suggest that duration is a correlate of stress in C & oacute;rdoba, as in other Spanish varieties, and small relative duration differences due to pretonic lengthening may be enhanced by a larger pitch change. Nevertheless, sporadically we also find cases where the pretonic is longer than the tonic. In these examples, pitch-accent alignment indicates the position of the lexically stressed syllable.
This paper presents a highly-granular syllable-contact analysis of Proto-French (2nd-9th centuries) stop epenthesis in minimally-rising sibilant or sonorant + sonorant clusters (cisera > cisdre 'cider', cumulu > cumble 'peak', camera > chambre 'room', molere > moldre 'grind-INF') within a parallel optimality-theoretic framework. Capitalizing on Split Margin Approach syllable-contact constraints, place agreement and faithfulness, and phonotactic constraints against [*tl, dl] sequences, the account offers a cohesive analysis explaining both the asymmetric gap in stop epenthesis for word-medial /sl/ clusters (i(n)sula > isle, *istle/*isdle/*iscle 'island') as well as for deviant clusters whose epenthetic velar stop defies the homorganicity requirement with the input coronal coda, such as medial /nl/ > [eta.gl] (spinula > espingle 'pin') and word-initial /#sl/ > [(e)skl] (slaitan > (e)sclater 'burst-INF'). The analysis elucidates these non-canonical reflexes, which have proven elusive in the literature, by drawing on other attested phenomena affecting early French minimally-rising clusters such as manner change (/tN, dN/ > [z.n], e.g. retina > resne 'rein'), regressive nasal assimilation (computare > conter 'count-INF'), repairs resulting from the phonotactic aversion to /tl, dl/ (Late Latin place change to [kl, gl], e.g. vetulus > ve[kl]us; Proto-French deletion, e.g. /tl, dl/ > [ll] spatula > espalle 'shoulder,' ustulare > usler 'burn-INF') and the onset of phrase-level prothesis of word-initial /#sC/ -> [#(e)sC] as of Late Latin, proposing a modified dating of /#sl/ > [#skl] and explaining how and why /sl/ and /nl/ flout the canonical stop epenthesis pattern.
The acoustic properties of two types of prosodic prominence, stress and focus at the lexical and sentence levels, respectively, are examined in Romanian. Real three-syllable words were produced by 9 native speakers in two sentential contexts: with focus on the target word (focus context) and with focus on a word later in the sentence (non-focus context). The position of stress was systematically varied, so it appeared on each of the three syllable positions. Analyses of over 3,500 vowels show that duration and F0 are robust correlates of lexical stress in Romanian, consistent with cross-linguistic patterns. Focus further enhances these stress-related properties, lengthening and raising the pitch of stressed syllables in particular, with smaller effects on unstressed ones. In addition, vowel space visualizations based on F1 and F2 indicate that stress produces slightly more peripheral realizations and that focus reinforces this tendency. These results demonstrate that stress and focus interact additively in Romanian, primarily through duration and F0, and they highlight how prosodic prominence is manifested in a Romance language that has been underrepresented in the literature.
In this paper I study two groups of verbalizations that can be formed out of the same adjectival base in Catalan (e.g. fosc 'dark'): derivative verbs, like fosquejar ('to be darkish, to go darker'), and parasynthetic verbs, like enfosquir ('to darken'). By providing a comprehensive analysis of these two groups of verbs according to which the former express either states or events of scalar change and the latter only express events of scalar change, I argue that parasynthesis does not hold in the first group because those verbs involve an impoverished morphosyntactic configuration in which there is no structural room for the prefix. This reduced configuration explains why these verbs, unlike their parasynthetic counterparts, have a particular semantics by which there is a degree that is slightly smaller than the adjectival standard of comparison, if they express a state, or by which there is an event of scalar change in which the increasing degree does not reach the adjectival standard of comparison, if they express an activity.
Valencian vowel harmony has been analyzed as a process that spreads the color features of the stressed open-mid vowels /'epsilon, '(sic)/ to a following final posttonic open vowel /a/ within the morphological word. Although this statement correctly depicts the prevailing pattern, recent experimental studies within the framework of laboratory phonology show that harmony may extend further. In light of these new data, we postulate that the scope of Valencian vowel harmony is conditioned by the prosodic structure, and that part of the variation found is due to the different ways in which clitics can prosodically integrate with their lexical hosts. We examine in detail the variation that arises from enclitics being parsed as either affixal or internal clitics and their interaction with the lexical host, providing additional evidence for the suitability of internally layered ternary feet. In our proposal, vowel harmony arises from licensing constraints demanding the features of the mid-open vowels to be attached to salient enough positions. We develop a prominence-based analysis in which features extend over increasingly larger prosodic domains in order to improve their perceptibility, giving rise to the whole typology of Valencian harmonic patterns without generating unattested forms.
While Standard Italian only displays word-final consonants in loanwords and some function words, Italo-Romance varieties show a wide range of variation in which major classes of consonants can be found in this position. This survey provides a first overview and finds that the northern varieties and the southern varieties are two different systems, the north gravitating to sonorants and the south showing a preference for obstruents. In the vast majority of cases, unattested consonant classes are excluded by non-application of apocope, which, in Optimality Theory, can only be accounted for with positional markedness or licensing, not with positional faithfulness. The typological relationship between different sets of final consonants as well as types of final clusters shows, furthermore, that an account of the southern system in terms of final onsets is undesirable.
Abstract The swarm -alternation shows two variants, where the locatum and location alternate their positions: in the locatum-subj variant ( The bees swarm in the garden ), the locatum appears in subject position, whereas the location surfaces as a PP. In turn, in the location-subj variant ( The garden swarms with bees ), the location appears as the subject, and the locatum is introduced by a preposition ( de/di “of” in Spanish, Catalan, and Italian). We argue that the locatum-PP is, in fact, the element introducing the initiator of the event in the latter variant. This is so because the subject is a non-theta selected element, functioning as the undergoer of the event; hence, it does not meet the requirements to initiate it. As to the locatum-subj variant, the locatum subject works as the initiator and undergoer of the event. Furthermore, we show that the argument structure of these verbs consists of both process and initiation phrases in Ramchand’s first phase syntax. While the location-subj variant is consistently unergative, the locatum-subj variant can also be unaccusative as is the case with verbs such as overflow or abound .
The swarm-alternation shows two variants, where the locatum and location alternate their positions: in the locatum-subj variant (The bees swarm in the garden), the locatum appears in subject position, whereas the location surfaces as a PP. In turn, in the location-subj variant (The garden swarms with bees), the location appears as the subject, and the locatum is introduced by a preposition (de/di "of" in Spanish, Catalan, and Italian). We argue that the locatum-PP is, in fact, the element introducing the initiator of the event in the latter variant. This is so because the subject is a non-theta selected element, functioning as the undergoer of the event; hence, it does not meet the requirements to initiate it. As to the locatum-subj variant, the locatum subject works as the initiator and undergoer of the event. Furthermore, we show that the argument structure of these verbs consists of both process and initiation phrases in Ramchand's first phase syntax. While the location-subj variant is consistently unergative, the locatum-subj variant can also be unaccusative as is the case with verbs such as overflow or abound.
This article presents the first analysis of the word order of Early Old Catalan by examining four texts dating from the 10th and 11th centuries. While the texts vary in nature, they all fall within the legal register, and they are all deemed suitable for the study of syntax based on extra-linguistic criteria. Approaching the data from a generative and cartographic perspective, it is shown that (i) broadly speaking, Early Old Catalan patterns with contemporary Romance texts in exhibiting an active low V2 grammar; (ii) it shares with other Old Romance varieties the use of expletive elements to satisfy the V2 parameter; (iii) Early Old Catalan differs notably from 13th century Catalan in the use of the left periphery and the application of V2. The findings in this study cast light over the fragmentation of the Romance linguistic continuum at syntactic level and open new avenues of research on the matter.
French wh-in-situ questions may be infelicitous when a quantifier or focus expression precedes the in-situ wh-phrase ('intervention effects'). However, these intervention effects are absent in a certain type of context. This paper offers an account of French wh-in-situ questions that explains the role of context. We propose that French wh-phrases can be interpreted in-situ by means of a contextually bound choice function. We further hypothesise that, due to an ongoing process of language change, there are two groups of French speakers. Both groups can interpret wh-in-situ via the choice function but one group may additionally use covert movement. We show how the choice function account along with this 'Two speaker groups hypothesis' accounts for the effect of context on intervention effects, as well as several other longstanding debates regarding the semantics and prosody of French wh-in-situ questions.
We investigate the status of weak crossover violations (WCO) in Spanish restrictive relative clauses. Claims in the literature lead to the expectation that Spanish does not show sensitivity to WCO in this context; however, ours is the first systematic investigation of the phenomenon in the language. We find that Spanish restrictive relative clauses do exhibit WCO violations: when the relative clause is headed by que, participants rate crossover configurations lower than configurations without crossover; when the relative clause is introduced by a quien, the status of WCO violations is less clear. We consider two explanations for our results in a quien clauses: potential structural differences between que and a quien, and the possibility for an appositive parse for some of our a quien configurations. Our results set the stage for further cross-linguistic investigations of crossover phenomena.
One major achievement in syntax has been a deep understanding of displacement in terms of Internal Merge. Therefore, displacement types initially resisting that analysis deserve scrutiny. This article investigates one. Latin verse permits tmesis – the division of “words” into nonadjacent pieces. In one subtype, radical tmesis, the cut is not obviously at a morpheme boundary. If it were not, radical tmesis would be theoretically recalcitrant. The article argues, however, that radical tmesis is actually derived by Internal Merge. The cut does occur at a morpheme boundary, despite appearances. Furthermore, the constituent orders radical tmesis produces can be derived syntactically, positing only independently motivated operations. Radical tmesis, then, is syntactic – supporting nonlexicalist frameworks, e.g., Morphology as Syntax. Even displacement types yielding apparently “irregular” outputs, then, can turn out on examination to be products of Internal Merge, a subcase of the elementary structure-building operation Merge – a theoretically welcome result, given minimalist aims.
This work presents original results from a new experimental approach to Differential Object Marking (DOM) using customized Acceptability Judgment Tasks (AJTs) in varieties of Spanish (Iberian, Mexican, Peruvian, and River Plate) and Portuguese (Brazilian, European), together with data from Catalan varieties obtained with the same methodology, Differential object Marking in Romance: The third wave , 279–314. Berlin & Boston: de Gruyter). We show that AJTs can be used to produce a fine-grained picture of similarities and contrasts in the variation of Ibero-Romance DOM along well-established prominence scales, such as the Extended Animacy Hierarchy. We also test a series of predictions from the literature for each variety. Although we find clear evidence for scalar effects, we question the explanatory potential of a finite set of categories on prominence scales, and we argue that in order to explain DOM in the varieties under investigation, the nature of such scales needs to be better understood.
Most analyses of wh -questions posit the existence of a dependency between the scope position and the theta position of the wh -constituent, both in ex situ questions and in in situ questions, but they disagree on the nature of this dependency and in particular on whether it is the same or different in the two strategies. The aim of this article is to verify how things stand in French, which is a mixed language allowing both in situ and ex situ wh -phrases in normal information seeking questions. We tested with two acceptability judgment tasks and two sentence-picture matching tasks whether in situ questions in French display the same subject advantage that has been widely attested in ex situ A-bar dependencies across languages. We found as expected a subject advantage in French wh -ex situ questions but an object advantage in French wh -in situ questions. We explain this seemingly divergent pattern by a) assuming that the same kind of movement dependency is established in both strategies but b) that its landing site is different and hence interacts differently with the clitic left dislocation structure (CLLD) we introduced in our experimental items. In both strategies, a processing advantage is observed when the wh -dependency and the clitic left dislocation structure (CLLD) dependency are nested, while crossing dependencies result in a cost in both cases.
This article confirms the existence in French of an overt morphological correlate to Merchant’s abstract E-feature, which I call special le . Taking as a point of departure my claim in previous work that special le is, in Modern French, the obligatory phonological realization of the E-feature in French predicate ellipsis, I undertake a close examination of a different type of ellipsis, namely, clausal ellipsis. I first show that, with the exception of sluices, both overt and covert versions of the E-feature in clausal ellipsis are available, although the choice between the two versions varies along register and construction specific lines. Second, I establish that there are, in French, two distinct types of clausal ellipsis. Type 1, exemplified by so-called modal ellipsis, only requires the pairing of an E-feature with a modal verb that takes a phasal complement. Type 2, exemplified by sluices and the ellipsis of the complement to a bridge verb, is instantiated by the pairing of an E-feature with the higher C-projection (C 1 ) of a CP-recursion structure licensed by the presence of a speech act. It is further argued that when C 1 is endowed with an E-feature, it must also be associated with an EPP feature, which is but one example of feature clustering among many others. A direct consequence of this is that type 2 clausal ellipsis requires, rather than allows, wh -extraction of a remnant in order to be licensed. Finally, it is shown that relative clauses that embed a bridge verb license type 2 clausal ellipsis only when they are of the ACD type. Regardless of any analysis, this observation entails that standard and ACD relatives undergo distinct derivations. I assume that, in order to avoid infinite regress, ACD relatives force a more complex/costly derivation: they must be late merged, and I argue that late merged ACD relatives contain a full copy of the DP relative head which constitutes, in the specifier of C 1 , the type of remnant that licenses type 2 clausal ellipsis; that is, a remnant merged with fully valued φ-features. In standard relatives, on the other hand, relative pronouns are minimal pronouns that bear unvalued φ-features that are valued in the course of the derivation via Agree between the head noun and the relative pronoun. As such, run-of-the-mill relative pronouns, having unvalued features when they transit through the specifier of C 1 , are not appropriate remnants for type 2 clausal ellipsis.
This article presents a modern theoretical approach to auxiliary selection in Standard European French. Assuming that have and be are allomorphs spelling out the functional head v Aux , it argues that the person feature identity of v Aux plays a central role in determining auxiliary choice: be is inserted when T and v Aux share the same person feature identity, and when identity differs, have obtains. Crucially, and unlike most analyses in the literature, this simple proposal provides a unified account of auxiliary selection for unaccusatives, transitives, and reflexive verbs. It explains why reflexive transitives pattern with unaccusatives, and as a result, the analysis challenges the view that auxiliary selection is sensitive to argument structure. A key assumption here is that have functions as the elsewhere auxiliary.
