LANGUAGE CONTACT AND LANGUAGE CHANGE IN THE FAETAR SPEECH COMMUNITY
A Proposal for a Dissertation in Linguistics in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
submitted by
Naomi Nagy
TABLE OF CONTENTS
In this dissertation, I analyze newly uncovered evidence of several linguistic changes which have taken place in the development of Faetar, an unwritten and virtually unstudied Francoprovençal dialect which has been spoken in a small farming village in Apulia, a province of southern Italy, for the past 600 years. Because the language has been in contact with Italian as well as the colloquial dialects of the area, it has undergone many changes, and is no longer mutually intelligible with Gallic Francoprovençal. However, unlike many dialects spoken in small isolated communities, Faetar does not appear to be succumbing to language death or shift. Unlike the situation found in most European countries in which local dialects flourished only until the institution of a national language, all the natives of Faeto still speak Faetar on a daily basis (reserving Italian for communication with outsiders), and most emigrants continue to speak Faetar for several generations after leaving Faeto. Faetar appears to be held in high regard by its speakers: they recognize a certain amount of prestige in this marker of their distinctness from other southern Italians. The question arises, however, of just how distinct the language is. In its centuries of contact with Italian and Apulian dialects, Faetar has changed in many ways, apparently adapting features of the languages with which it is in contact. These diachronic changes are the topic of my dissertation. I explore several changes, showing the effects of Italian and Apulian on the phonology, morphology, and lexicon of Faetar. It has been demonstrated that variation in apparent time (synchronic variation) corresponds to a diachronic change in progress, especially when accompanied by social and style stratification (Guy, Wikle, Tillery, and Sand 1993, Labov et al 1972:6). I put to practice this established link between synchronic and diachronic variation in order to explore the changes which have taken place during the centuries of contact between Faetar and standard Italian and Apulian dialects. The patterns of synchronic variation which I present can be interpreted as evidence of sound change. In addition to adding a new description of a language to the literature, these analyses provide test data for frameworks which have been posited concerning the relationship between sources and effects of contact-induced language changes. I examine claims made in recent works by Thomason and Kaufman (1988), Van Coetsem (1988), and Guy (1990) to see if they adequately account for the linguistic change which have taken place in Faetar. Each of these authors suggest that certain types of linguistic elements are more easily borrowed than others, and that the kinds of transmission which may occur depend on the social relations between the speakers of each language. However, quantitative evidence is not provided by any of the authors to support their claims about the relative frequency of different types of change. Nor is an in-depth study of any one language contact situation presented in the above cited works. I propose to address these gaps in the literature by presenting my detailed examination of the relative degrees of lexical, phonologic, and morphologic influence on Faetar by other Italian dialects, and accompanying this discussion by an explicit description of the contact situation in Apulia. In the process, I develop a reproducible method for comparing degrees of linguistic change across linguistic levels. Linguistic change has traditionally been analyzed using diachronic evidence only, beginning with the Neogrammarians' family tree model. Later proposals, stemming from the wave model (Schmidt 1872), highlight the importance of considering synchronic evidence as well (Hock 1991:450-5). Variation across dialects or speakers has been considered an indicator of linguistic change by linguists such as Hermann (1929) Sommerfelt (1930), and Reichstein (1930). The most far-reaching analyses of this type are those of Labov (cf. e.g. 1963, 1966), who designed a method of investigation for analyzing interspeaker variation in vernacular speech to determine what types of sound changes are occurring in the community. He states:
The methods of our study involve the further claim that linguistic theory must take social factors into account for a rational account of language change. Within the theoretical model of a homogeneous speech community there are serious problems of conceptualizing change, and it is not surprising that those who wrote on historical linguistics within this framework rejected on a priori grounds the possibility of studying sound change in progress (Bloomfield 1933:365; Hockett 1958:457; Chomsky 1965:3). When we abandon the identification of structure with homogeneity it is possibly to construct a rational model of change and begin to observe it. (Labov et al 1972:3).
The use of methods developed within the field of sociolinguistics for analyzing synchronic variation have added to the practice of historical linguistics. Synchronic data can be examined thoroughly and systematically, using large data sets, whereas traditional diachronic analysis involves a lot of guesswork about dead languages for which there is limited documentation. Synchronic evidence, in oral form, is abundantly available to anyone willing to do the fieldwork and can provide answers to many problems which must remain unsolved when only a limited amount of old evidence, in written form, is available. The sociolinguistic method of analyzing synchronic variation also has the advantage of being able to explore patterns which vary according to style and extralinguistic factors such as speaker's social status, while a traditional historical analysis can only make use of written forms, which are, usually, in the formal style of the higher classes of speakers. I discuss several linguistic changes which distinguish Faetar from its Gallic Francoprovençal relatives. For each change, I describe the variable and the envelope of variation, present a formal analysis of the pattern within a variationist framework, and propose both synchronic and diachronic sources for the change. Because I use information from current sources, I can produce a more realistic description of the linguistic history, making use of acoustic evidence rather than guessing at pronunciation from the orthography. My analysis is conducted using the interview, elicitation, and quantitative analysis techniques developed for the study of synchronic variation by sociolinguists since Labov (1968). I use a corpus of approximately 50 hours of recorded speech from more than 60 speakers, which I recorded over the past 3 years. The first change to be examined is the appearance of geminates in Faetar. Gallic Francoprovençal had lost all consonantal length distinctions by the fourteenth century (Bourciez 1930:305), when Faeto was settled, so this phenomenon appears to be an innovation due to contact with Italian, a language which has geminate consonants, and which I have shown to be a frequent source for lexical borrowing (Nagy 1993). However, I show that the pattern of gemination in Faetar is not identical to that in standard Italian, suggesting that there may be other influences at work as well. I also examine the possibility of an internally motivated sound change: one which can be explained with reference only to Francoprovençal (Nagy 1994). The second change is a process of variable end-of-word deletion. Two possible sources for this process present themselves immediately. Gallic dialects exhibit a large degree of diachronic final segment deletion when compared to their source language, Latin. Colloquial Italian has a synchronic process of final syllable deletion. Careful analysis of the three systems (Gallic, Italian, and Faetar) will indicate which sources are responsible for the variable pattern in Faetar. For each of these linguistic changes, I conduct a series of experiments to quantitatively describe their phonetics, phonology, and morphology, and discuss the relative merits of each historic source proposed for them. In this, I follow Labov, who promotes investigations which begin with phonetic results, rather than an abstract theoretical position (Labov et al 1972:269). A detailed description of these phenomena will augment the scarce data available for this unwritten language variety. This will be a useful contribution to the reconstruction of Francoprovençal as it was spoken some 600 years ago, long before enforced standardization in France nearly wiped the language out. The structure of this dissertation will be as follows: Chapter 1 provides a history of Faeto and Faetar, and an ethnographic description of the speech community. This chapter also contains an account of the status of nonstandard languages in Italy, and discusses the use of nonstandard languages as teaching languages, a proposal which is currently being considered by members of the Faetar speech community. Chapter 2 describes the fieldwork and research methods which I used in collecting the data. Chapter 3 will give a brief phonologic description of Faetar, comparing and contrasting Faetar's system to those of Italian and Apulian dialects. Chapters 4, 5, and 6 contain analyses of the linguistic changes described above and include descriptions of the current forms in the language; descriptions of the corresponding forms found in Italian and Francoprovençal, the two source languages; a formal analysis for each change; a description of how the changes took place; and a discussion of how these processes can be formally accounted for in a variationist framework. These chapters also contain a critique of previous accounts of Faetar, focusing on the one substantial account, Das Frankoprovenzalische in Süditalien. Studien zur synchronischen und diachronischen Dialektologie by Dieter Kattenbusch, which contains a detailed description of the language as it was spoken in the 1970's. Chapter 7 contains a discussion of the literature regarding the relationship between types of social contact and the ensuing linguistic changes. Also in this chapter, I propose a method of quantifying linguistic change at different levels of the language. Chapter 8 summarizes my findings and show how they all fit together in a quantified description of linguistic change reached by a combination of diachronic and synchronic analyses. In this proposal, I have written a section corresponding to each of the proposed chapters listed above. There will also be an expansion of the introduction, in which I review the relevant literature on methods of studying linguistic change, contrasting the traditional method of examining only diachronic evidence and more recent attempts that incorporate synchronic evidence.
1. Description of the speech community
This chapter presents an overview of historic sources which discuss the settlement and history of Faeto with an attempt to reconcile the historical and linguistic data. This includes information that I have gathered from written historical accounts of Faeto's history, from the parish and city hall records, and from a compilation of census data edited by the Apulian regional association (Regione Puglia 1989). I also present ethnographic information which I have gathered from the inhabitants of Faeto, as much of the official historical sources have been destroyed. This information is highly pertinent in the analysis of sources of linguistic change, as degree and type of contact have been shown to play an important role in language change (Thomason and Kaufman 1988).
1.1 History of Faeto and Faetar
A review of available literature on the settlement of Faeto suggests that the town was founded in the late fourteenth century by a small group of people from the region of France between Lyon and the Alps, a region distinguished by its dialect, Francoprovençal (DeSalvio 1918:45-46). There is some confusion in the received history of Faeto: its origin is attributed to a land grant from King Charles d'Anjou to his soldiers upon their victory at Lucera over Frederick II. However, Charles d'Anjou's reign began after Frederick II died. A second commonly expressed hypothesis is that the town was founded by Waldensians who fled France due to religious persecution in the fourteenth century (Melillo 1981:39). Further archival research may bring forth more accurate information about Faeto's origins and early history. To supplement the scarce historical records, I have conducted a preliminary dialect geography analysis of the region of France from which Faetar comes (Nagy 1992a). In this work, I compare the relative number of cognate words in a number of different Francoprovençal dialects to Faetar. I will use the method I developed on an expanded set of cognates, to produce more conclusive results, in hopes of narrowing down the point of origin of the current speakers of Faetar. This will allow for more accurate comparison of Faetar to its ancestral language family and expand on the comparative work initiated in Morosi (1890).
This section presents an ethnographic description of Faeto, discussing issues such as the degree of Italian-Faetar bilingualism, how long bilingualism has been the norm in the village, and what languages, aside from standard Italian, may influence Faetar speakers. This information has been collected in interviews which I conducted as a participant observer. It includes speakers' reports on when and for what purposes Faetar, as opposed to Italian, is spoken, and with whom. My finding is that all adult members of the community, except recent spouses from other towns, speak Faetar on a daily basis. Many children do not speak the language, although they understand it. It appears that teenagers regain an interest in speaking the language, coincidentally (?) at the age at which they must go to a neighboring town for school. Faetar is used in the home, on the street, in the town hall, and in stores and bars. The only place where I noted its absence is in the church, as the priest is not from Faeto, and, even there, the choir sings some songs in Faetar. School is conducted in Italian, with some research projects on Faetar as part of the curriculum for 10 and 11 year olds. The students interview their elders in Faetar and, with the assistance of the teachers, transcribe and translate the stories told to them. Children have reported that their teachers do sometimes use Faetar in school, but I have not made first-hand observation, as the school is closed during the summer, when I did my field work. All members of the community that I have met are fluent in Italian, in addition to Faetar. It has been reported that some of the older farmers who live outside the village do not speak much Italian, but I haven't yet interviewed anyone who did not speak Italian. Replies to my questions about the history of bilingualism in Faeto range from "it's just happened in the last few generation" to "it was necessary since the beginning." I hope to find more reliable information on this subject in future research.
1.3 The status of "dialect" in Italy
I summarize my discussions with a number of Italians, from throughout the country, which indicate that "dialect," or non-standard languages spoken in informal situations, are popularly considered to be corruptions of Italian, rather than alternative languages which are appropriate in certain contexts. I compare and contrast views on Faetar to this general opinion of "dialect." Faetar does not suffer the same stigmatization as many other dialects, due to its distinct origins. Its speakers, in fact, are quite proud of this distinguishing characteristic. Faetar cannot be considered merely a variant of standard Italian, as there are abundant indicators of the Francoprovençal origin of its grammar. This issue is important for the analysis of the linguistic changes in the language, as speakers of a dialectal variant are often perceived quite differently from speakers of a different language. Faetar represents a situation not often mentioned in typologies of contact situations: a minority language with relatively high prestige. Because Faetar clearly has origins which differ from those of the surrounding regional dialects, it is seen as a marker of unique social identity, distinguishing its speakers from other Italians, at least according to its speakers. Comments from my interviews with members of other speech communities in the regions reveal how they view the language: as an anomaly, but not entirely unexpected in southern Italy. This factor must be considered in determining the type of influence which Italian has on Faetar, since, unlike most language contact studies, this one does not involve a contact situation where the non-standard language has lower prestige than the standard. Many speakers have suggested that I design an orthography for them and write a grammar book, two prerequisites for using Faetar as a teaching language in the local school, according to Italian law (Melillo 1991). This oft-repeated request serves as further evidence of the recognition of Faetar as a language in its own right, in the eyes of its speakers, and the level of prestige with which it is regarded. These views are important to consider in evaluating the effect of contact situations on linguistic change.
Because Faetar is not an institutionalized or even written language, and because its speakers are so isolated, it has not been the subject of much linguistic research. Several people have written brief descriptions of the language, including Kattenbusch (1982), Morosi (1890), Melillo (1956-7), and De Salvio (1918). With the exception of Kattenbusch's work, these consist solely of catalogues of words and phonemes, with little linguistic analysis. Kattenbusch provides some phonemic analysis, but his claims and his data are sometimes contradictory. Because this language is in danger of disappearing from the province within a few generations, as the town is quickly decreasing in population, it is important that an accurate description of the language be preserved, accompanied by recordings, for further analysis. My research will aid in accomplishing these goals, providing a description which goes further than a simple catalogue of words and sounds, as well as a large corpus of recordings and transcriptions. This corpus has been compiled through participant-observation fieldwork over the past three years. This fieldwork has four major components. First, while living in Faeto, I have been able to observe first-hand how, when, and by whom Faetar is spoken, and to note that actual behavior does not always follow reported behavior. A discussion of the patterns of language usage is included, as support for claims about the relative prestige of Faetar and Italian and to show what type of language contact situations arise in and around Faeto. Second, these observations are supplemented by tape-recorded interviews in which I asked speakers a series of questions about life in Faeto, including language usage. I have obtained approximately 30 hours of recorded conversational speech from 40 informants. These recordings contain some one-on-one interviews, but also a number of small-group conversations, in which the speakers are talking to other native speakers of Faetar, in a more natural and quite casual style. In addition to the body of natural speech data which I recorded, I have collected recordings and transcriptions of three interviews made in the 1970's by Dieter Kattenbusch, of two made by Anna Fragasso in 1977, and of one made by Carmela De Fino in 1970. This set of recordings serves two purposes: in addition to gathering information about the speakers' views on language usage, it provides a large body of spoken text which can be analyzed to determine what patterns exist in the language. Such a large body of naturally-occurring speech is necessary for analyzing linguistic variation. The third type of data was collected using a controlled-topic elicitation task, in order to get easily comparable results from all speakers. For this task, 30 men and 30 women, evenly distributed across age-groups from 6 to 87 years, were each asked to describe pictures in a children's story book and name the objects shown. This will be discussed further in the section on initial geminates, as the task was designed to show the effects of context on geminate pronunciation. The fourth type of data collected is the results from a series of word-list elicitation tasks. This provides immediately comparable data from 15 speakers for pronunciation of 18 minimal pairs of words, for the purpose of analyzing geminates. Several repetitions of each word were recorded, in a carrier phrase, in randomized order, with elicitation being accomplished by flash cards illustrating each word. The recording have been digitized and the lengths of the pertinent segments measured using Xwaves, an interactive acoustic display package.
This is a descriptive chapter, provided as a back-drop for the analysis of phonologic changes discussed in chapter 4, 5, and 6. I present the phonemic and phonetic inventories, syllable structure patterns, and stress assignment rules for Faetar, standard Italian, and Apulian dialects.
3.1 Phonemic and phonetic inventories
In this section, I draw on previous descriptions of Faetar (Minichelli (1994), Melillo (1956-7), Morosi (1892), Rohlfs (1973), and particularly Kattenbusch (1982)), correcting them where my own observations require, in order to provide phonemic and phonetic inventories of the language. I also describe the contexts of allophonic variation for each phoneme. Using recordings from speakers in neighboring towns, as well as descriptions from Valente (1975), I present the phonology of Apulian dialects for comparison. A phonology of standard Italian is adapted from Nespor (19__).
Virtually all possible syllable structures are found in the language, from V to CCVC (word initially) and CVCC (word finally) (Nagy 1992). Long vowels and diphthongs are also attested; in our data, for example, we have bimoraic nuclei consisting of [i:], [ei], and [ai]. Certain vowels in unstressed syllables regularly surface as schwa. (Nagy 1992b). Examples of words with each type of syllable structure will be provided.
3.3
Stress
The stress system of Faetar is based on quantity-sensitive left-headed binary
feet, or moraic trochees, with main stress on the rightmost foot. Such feet
are formed from left to right, with the result that unfooted syllables (i.e.,
degenerate feet) may appear at the right edge of a word (and occasionally
word-initially or -internally as well). For purposes of stress, syllables with
a rhyme consisting of V: or VC (or longer) are considered heavy (Reynolds and
Nagy 1994). A number of examples will be given to support this analysis.
Additionally, a brief discussion of words which do not fit the above
description will be included.
Geminates in Faetar
Previous research has indicated that Faetar has geminates of two types: lexically marked geminates in medial position and phonologically triggered geminates in initial position (Kattenbusch 1988:169). However, I show that this description is not completely correct. Phonetic data is presented to show that there is, indeed, distinctive consonant length in medial position, but not in initial position. This finding is somewhat surprising, as all transcriptions of Faetar, by native speakers and outsider linguists alike, include initial double consonants. Because the geminates in medial and initial position differ considerably, I consider them separately, in the following two chapters.
The presence of geminates in Faetar is unexpected, as Gallic Francoprovençal did not have any geminates by the fourteenth century, when the first settlers of Faeto came to southern Italy. All Latin clusters had been simplified in Gaul by that time (P. Lloyd, K. McMahon, P. Léon pc). According to Bourciez, geminates had been simplified throughout Gaul by the ninth century, with the exception of the trilled [RR] (1930:305). Geminates in Faetar have been discussed in Dieter Kattenbusch's detailed description of the language, Das Francoprovenzalische in Suditalien (1982). He indicates that geminates are distinctive in medial and final position, as the following minimal pairs show, but provides no phonetic evidence (Kattenbusch 1982:168).
4.1 Proof of existence of medial geminates
In this section, I describe a set of experiments which show that Faetar does have a length distinction in medial consonants, rather than free variation in consonant length, or predictability based on environment. In the first experiment, a comparison was made of the durations of medial consonants which had been transcribed as single and geminate in interviews conducted and recorded by Kattenbusch. Because a lot of the variation observed in this experiment may be due to variation in speech rate, a second, more carefully controlled experiment was designed. Minimal pairs of words were selected and each word was elicited in a carrier phrase, six times, in random order, by showing pictures of the objects to the speaker. Ten speakers were recorded for this experiment. I show that the results of both experiments indicate quite clearly that there is a length distinction in medial-position consonants in Faetar. This supports previous analyses of a medial consonantal length distinction in Faetar, providing a firm departure point for an analysis of the historical source(s) of this change. Further experiments will study speakers of varying ages and social status in order to determine whether this is an ongoing change or one which has reached completion, applying principles of apparent time's relation to real time (Labov et al 1972: 6).
I next discuss possible explanations for the presence of geminates in Faetar. There are three possibilities. The first is an internal change. For that, a pattern in Francoprovençal must be posited as the source for a regular sound change rule. The second is lexical borrowing, with Italian and other local dialects as the external source. The third is also an external change, where contact with Italian or other local dialects causes the introduction of a new distinctive feature (or a new permissible structure), a form of grammatical borrowing. Francoprovençal, standard Italian and colloquial Italian are examined for possible source patterns for the medial geminates in Faetar. The proposed differences in types of linguistic change effected by internal and external sources will be discussed in the final chapter.
In order to consider the first possibility, an internally-caused change, a description of the Francoprovençal system is necessary. All accounts indicate that Gallic dialects had lost the consonantal length distinction (which was present in Latin) by the 14th century, when Faeto was settled. I consider the possibility that there was some other pattern in Francoprovençal which was systematically realized as distinctive consonant length when the language came into contact with Italian. Several possibilities of sound patterns which could change to geminates exist. The first is consonant clusters in Francoprovençal which may have undergone assimilation in Faetar. The second is that the deletion of a vowel in Francoprovençal would form a consonant cluster in Faetar, which could then undergo assimilation. The third is that a distinction in vowel length could be transformed into a distinction in consonant length. In order to see if any of these patterns exist, I am compiling a list of Francoprovençal cognates to Faetar words from A. Duraffour's Glossaire des patois francoprovençaux and am conducting a comparison to see if there is any possible pattern in Francoprovençal which could, by regular sound change rule, perhaps accompanied by analogy, be realized as a consonant length distinction in Faetar.
If the geminates in Faetar cannot be traced to Francoprovençal, a solution from Italian must be sought. In order to determine exactly how Italian influenced Faetar gemination, an examination of Italian cognates is necessary. If it can be shown that Italian words have geminates exactly where Faetar words do, then some form of lexical borrowing would be evident. However, borrowing from Italian to Faetar is not superficially apparent in many cases. Here are just a few examples of Faetar words containing geminates where one does not find corresponding geminates in the Italian words of the same meaning.
Faetar Italian English [êarrier\] via, strada street [kwattra] ragazzo boy [k\cciy] addormentarsi to go to bed [accetunt] comprare to buy [arriy] ancora again
As close cognates cannot be found for all of these words, a different solution must be sought: nonstandard dialects spoken around Faeto. A number of dialect atlases as well as the periodical Lingua e storia in Puglia will be searched for possible sources of the Faetar words containing geminates. Additionally, questions about these words have been incorporated into the interviews that I conducted with inhabitants of the villages neighboring Faeto.
4.3 Analysis: borrowing of a phonological structure
Because it is clear that the entire pattern of gemination in Faetar cannot be accounted for by lexical borrowing from Italian dialects (some of the words have clear Francoprovençal cognates but not Italian ones), a different explanation is necessary. I propose that Faetar has borrowed the feature [length] by adopting the Italian structure permitting linking of two timing units to a single segment. Thus, length has been incorporated as a distinctive feature in Faetar phonology. This type of linguistic change is not unknown, Thomason and Kaufman provide examples of borrowing of a feature or distinction which did not exist in the base language, but indicate that it is a relatively rare event (Thomason and Kaufman 1988:85). Kattenbusch analyzes long consonants as two utterances of the single consonant phoneme, adopting the "biphonemic solution" of Muljacic'. (Kattenbusch 1988:169-70) This analysis violates the Obligatory Contour Principle, by allowing two identical segments adjacent to each other on the segmental tier. Kattenbusch must stipulate that, in these cases, the two consonants share a single closure and release, while, normally, every consonant has its own closure and release. Theoretically, this appears identical to an analysis in which a geminate consists of a single segment linked to two timing slots. Analysis of the acoustic displays of geminate and single consonant production (spectrograms and time waveforms) will be conducted to see if this analysis is appropriate. Reference to phonetic investigations of geminates by Han (1992), Basboll (1989), Lahiri and Hankamer (1988), and Bertinetto (1974) will aid in constructing this investigation.
5. Initial geminates in Faetar
Although no speakers of Faetar are literate in the language, there have been sporadic attempts to write songs, signposts and short stories in Faetar, always accompanied by an Italian translation. One common thread in each of the authors' unique transcription systems is the presence of many initial double consonants. The authors are not consistent in their use of the double vs. single consonant: certain words appear transcribed both ways, and phonologic environment does not predict whether a single or double consonant will appear. Kattenbusch, one of the most systematic analysts of Faetar, makes the claim that the short and long forms of obstruents in initial position are non-distinctive allophones of the same phoneme (Kattenbusch 1988:169). He schematizes their predictable distribution in the following chart:
| after pause after vowel after consonant However, many pairs of words listed in his glossary suggest that there is a phonemic distinction of consonant length, even in initial position. The following chart shows a few such pairs from his glossary. chart omitted for lack of IPA symbols on WWW In a preliminary elicitation task with one informant, no distinction in the initial sounds for these words was noted (by us or by the informant) for such forms. One explanation is that the above citation forms are different in Kattenbusch's glossary due to their differing syntactic environments in his recorded speech. However, the last two are cases of both transcribed geminate and single consonants following identical definite article, e.g. [la ttan\] and [la tant\], [la mmoßk] and [la mon\k\]. Finding both forms in identical environments suggests that Kattenbusch's description of the distribution being solely determined by preceding segment is not correct, if his transcription is accurate. Furthermore, even his transcriptions of recorded interviews do not consistently follow the distribution which he laid out in the above chart. Prior to my investigation, the description of Faetar was: initial geminates appear, but not systematically, in all written forms of the language. No one has determined exactly where or why they appear, nor what their source is. To try to give some rationale to this aspect of the phonological system of Faetar, I designed a series of experiments to shed light on the following questions:
1) Is there any phonetic length distinction for consonants in Faetar? 2) If there is any difference, is it phonemic? That is, is it lexically marked? 3) Is it morphosyntactically triggered? If so, is gemination triggered by the same environments as in Italian?
In addition to improving the description of the language, I explore possible historical sources to explain the sound change which occurred in the evolution from geminate-less Francoprovençal to geminate-ful Faetar. This adds the question:
4) Can a historical explanation be provided? Is there some pattern either in Francoprovençal or in Italian which surfaces systematically as a distinction in consonant length in Faetar?
5.1 (Lack of) evidence of the existence of initial geminates
Using the medial position length distinction which was illustrated in the section on medial geminates as a baseline, several experiments were conducted to see if a length distinction is maintained in initial position. The first was an experiment using a controlled environment for elicitation, to be sure that no factors, other than potential lexical marking, were causing length variation. Six repetitions of 12 minimal pairs of words were elicited in random order, by six speakers. In a second experiment, Fosler and I selected minimal pairs of words from recorded and transcribed texts in Kattenbusch's book. Durations of transcribed initial double consonants were compared to corresponding transcribed single consonants. These data show that there is not a consistent lexically marked length distinction for any of the consonants recorded. Having concluded that Kattenbusch's transcriptions of lexical initial gemination in his glossary are in error, the next step is to examine contextual effects, in an effort to find a source for the common decision to transcribe frequent initial geminates. I turn now to a consideration of Kattenbusch's proposed phonological rules. His claim that geminates are found in initial position following vowels is not supported by the duration data. Although the post-pausal and post-consonantal consonants measured in the second experiment have consistently short durations, always less than 100 milliseconds, and usually less than 50 milliseconds, the putative geminates have a broad distribution, spanning from 20 to 130 milliseconds. The difference in mean duration of the two groups is not significant. This suggests that geminates have the property of variable length, while single consonants are consistently short, if we are to accept Kattenbusch's transcriptions of the words. Because geminates are normally categorized as being longer than single consonants, rather than being of such variable length, one must consider the possibility that the segments which we are calling geminates, but which are short, differ in some way from the long geminates. I consider several factors which may be responsible for the length variation in the transcribed geminates. The first is that the length of the following vowel affects the length of the initial consonant. Because there are several instances of identical words where some tokens have long initial consonants and others have short ones, this is not a feasible explanation. For example, the word [ppa] ('not') was uttered with varying initial consonant lengths ranging from 26 to 132 milliseconds. These measurements span the range of durations measured in the second experiment. Neither the phonetics of the previous word nor the syntactic relation between the words can be a factor, as is seen in examining a subset of the previous data. If we consider the duration of [pp] only in the cases where it is uttered in one particular environment, we still find a very broad range. The durations of [pp] in the phrase [e ppa] (`is not') are 26, 30, and 122 milliseconds. Because the above experiments were based on a small data set, with very few cases which allowed for direct comparison of words, another experiment was devised which, while still eliciting free-flowing, natural speech, controls the topic enough to cause much repetition of each word. For this task, a children's picture book was used. (Amery and Cartwright 1987) Each page contains a drawing of a scene. Several of the objects in the scene are also shown in isolated pictures at the bottom of the page. Speakers were asked to say the name of each object in isolation and then to describe the scene. This allowed for the elicitation of a very careful pronunciation of each word in isolation and then several repetitions of the word in sentence context. Sixty speakers, selected to provide a sample balanced for age (ranging from age 6 to 87) and sex, and ranging across social classes, were recorded performing this task. This will allow for analysis of variation due to social factors. The duration of initial consonants will be measured and mean durations compared for a number of different internal factors, including preceding segment (whether it is a consonant, full vowel or schwa), syntactic relatedness of preceding word, lexical marking, and following segment. A preliminary analysis of the internal factors for just two phonemes, as uttered by one speaker, indicate that there is no effect of preceding consonant, syntactic relation, lexical marking, or following segment. There is a great deal of variability in initial consonant length, even in controlled environments. The preliminary results contradict an analysis likening the process to raddoppiamento sintattico, a process in Italian which causes gemination of the initial consonant in specific morphosyntactic environments. In Italian, certain proclitics and oxytonic words trigger initial gemination of the following word. An example is the phrase caffè ffreddo' cold coffee,' in which the initial consonant of the second word is lengthened (as compared to the /f/ in aqua freddo `cold water') because it follows a tautophrasal word ending in a stressed vowel. Although Faetar appears to have borrowed the feature of distinctive length and can use it in lexical marking, it does not extend to the full use of distinctive length in Italian. That is, the length distinction is not used for morphosyntactic purposes, as in Italian raddoppiamento sintattico. This supports claims that lexical borrowings precede grammatical borrowings in contact situations of this sort (Thomason and Kaufman 1988:37). The current state of the language also supports Bybee Hooper's claim that `genuine phonological rules' can't be borrowed (Thomason and Kaufman 1988:16). Once the role of linguistic factors in determining initial consonant length have been understood, external factors can be considered and an analysis of interspeaker variation performed. The external factors to be considered include sex, age, social status, elicitation type (word list vs. narrative), and addressee (me vs. a native speaker), as all of these have been shown to cause intraspeaker variation (Labov 1972, Rickford and McNair-Knox 1992 inter alia).
Before abandoning the possibility of initial gemination in Faetar, a point to consider is that length is not the feature responsible for the perception of initial geminates. There may be some other process, a type of fortition, which is perceived as distinguishing certain consonants. This could include a difference in manner of articulation, voicing, frequency, or syllable timing. A two-step process, currently halfway complete, is necessary to see if this is the case. The first step is to determine whether native speakers actually hear this distinction by a controlled perception experiment in which the initial consonant is the only possible clue to the identity of the words. For this purpose, I have conducted a commutation task with nine native speaker subjects which show that there is a consistently perceived difference between pairs of words with the purported initial length distinction. The second step, which still needs to be taken, is to conduct an acoustic analysis to determine what cue is being used to make the distinction. Factors to be considered are voice onset time, relative amplitude of the consonant, and relative pitch of the following vowel. Several works will be consulted in designing the analysis technique, including Lahiri and Hankamer (1988), Han (1992), Hambert and Elugbe (19??) and Abramson (19??).
As with the medial geminates, analysis of the possible source languages is necessary to find an origin for the initial geminates. Standard Italian, regional Italian dialects, and Francoprovençal will be compared to Faetar to see if there is any pattern which, by a regular rule (and analogy), can account for the perception (if not production-- which remains to be seen) of some distinction in initial consonant length or strength.
5.3.1 Italian initial geminates
In addition to having lexical geminates, Italian has the process of raddoppiamento sintattico, which causes gemination of the initial consonant in specific morphosyntactic environments (discussed in 4.1). To determine whether the same patterns exist in Italian and Faetar, the first step will be an acoustic analysis of Italian initial geminates in the speech of Faetar speakers. If reliable acoustic cues can be found, a comparison with the Faetar initial consonants will be made, to see if their production is identical.
5.3.2 Francoprovençal initial geminates
Francoprovençal has no pattern of initial gemination. Nor is there any pattern, such as initial clusters or long vowels, which has been transformed into initial geminates in Faetar (Nagy 1994).
5.4 The Phonology of geminates and single consonants
In addition to using acoustic analysis, I will examine the phonologic behavior of geminate consonants to see if it differs from that of single consonants. One case to examine will be the allophonic variation of certain consonants. /[[integral]]/ has the following allophones: [[[integral]],v,w], /t/ has [t,d], /s/ has [s,z,ß] , /ß/ has [ß,[[Omega]]], /n/ has [n,~], and /k/ has [k,g,(c)]. Kattenbusch describes complementary environments for some of these allophonic sets, and free variation for others (Kattenbusch 1982:154-164). If it can be shown that the geminate forms of these consonants behave differently from the single forms (prediction: they do not undergo the voicing and frication changes because their structures do not exactly meet the requirements of the rule), it will be further evidence of the phonemic status of geminates. A second issue which may support the phonemic status of geminates is the relation between consonant length and vowel length. Kattenbusch states that geminates may only follow short vowels (Kattenbusch 1988:168-9). The large body of recorded speech which I have collected, and particularly the word list elicitations, provides useful data for testing this claim acoustically. If true, it will counter the possibility of geminates having phonemic status, as their description will be fully predictable (unless, of course, it can be shown that vowel length is determined by following consonant). A third issue involves the variable presence of word-final schwa. It may be that schwa is deleted in cases where syllabification is possible without it and retained when necessary to avoid certain types of consonant clusters. My hypothesis is that a schwa will appear between a consonant-final word and a geminate-initial one to prevent triconsonantal clusters, but not between a consonant-final word and a single consonant-initial word, as such two-phoneme clusters are permissible (cf. Kattenbusch (1982:173-175) for a list of licit clusters). The controlled elicitation task which I conducted provides the necessary evidence to confirm or deny this trend. Examining each of these cases in light of the new analysis of initial geminates will greatly enhance the description of the language, making it possible to provide a prosodic analysis with consistent descriptions of syllabification, apocope, and stress assignment processes. I have made a preliminary attempt at a prosodic description of Faetar (Nagy 1992b), but this can be much improved upon, once the question of geminates in initial and final position and the existence of a phonological word-final deletion rule (discussed below) have been assessed.
The third phenomenon to be examined in this dissertation is a word-final deletion process. Certain words have variable realizations, such as the Faetar word for Naples, which can be pronounced [na], [nap], or [napp\], the word for family, variably realized as [fami], [famiÒ], or [famiÒ\], and the city name Foggia, which is realized as [føÊ] or [føÊÊ\]. I describe the process, within the framework of Optimality Theory, investigating the limitations on when deletion can occur and the variability of the phenomenon. This is a necessary step before comparing this pattern to patterns found in other Italian and Gallic dialects, in order to determine the source of the process in Faetar.
The first task will be to determine the envelope of variation for this process. This requires considering both internal and external factors. An examination of my recordings of word list elicitations and natural speech will allow me to answer the following questions about both linguistic and extralinguistic factors:
1) Is this phenomenon restricted to certain words? 2) Is this phenomenon restricted to certain phonemes? 3) What prosodic unit(s) is/are deleted? 4) Is deletion restricted by number of segments or by syllabification requirements? 5) What role does word stress play? 6) Are there environments where deletion is mandatory? 7) Do all speakers exhibit this deletion process? 8) Does the speech style (casual vs. formal) affect the rate of deletion?
6.2
Phonological rule or fast speech process?
The next question to address is whether this process is a language-specific
phonological rule or is the result of universal fast speech processes, in which
case a historical explanation is unnecessary. This issue will be addressed in
several ways.
First, by examining the realizations of the words in question, I will
determine whether this process is gradient or has a bimodal distribution. That
is, are the deleted forms simply the most extreme stage of word-final
(post-stress) weakening which occurs in all speech, or must the final segments
be fully realized whenever they are not deleted, suggesting that there is a
phonological rule which may or may not apply?
Second, I will look at effects of morphological boundary. Morphology has been
shown to affect deletion rate in certain processes of English, Spanish, and
Portuguese (cf., Labov et al. 1968, Guy 1991, Hochberg 1986 inter
alia). If this is the case in Faetar, it will be evidence that the
deletion is a phonological rule, as lower-level rules should not interact with
the morphology. A quantitative analysis of deletion rate according to
morphological environment will be made. This will provide a description of the
synchronic variation, which can then be compared to possible historical sources
of the deletion process.
6.3 Sources of the change
There are two possible sources of the Faetar deletion rule, if it is indeed a phonological rule. Old French underwent a process of post-tonic vowel deletion in its derivation from Latin (Schwan-Behrens 1932:58-63). Such patterns may be continuing in Faetar, perhaps by a process of analogy. Although usually historical explanations should not be posited for synchronic change, as speakers are unaware of the history of their language, speakers who are Faetar-Italian bilingual have Italian (which did not undergo the deletion process) available for comparison and, thus, may be "conscious" of the vowel-deletion which has occurred in Gallic dialects. The second possible source is a current synchronic process in Italian, where word-final segments are variably deleted. For example, in southern Italian dialects, the words ricotta, pasta, and spaghetti are variably realized as [r^køtt], [past] and [ßpag'tt], having undergone final vowel deletion. A brief review of the literature on these two phenomena will be presented, and I shall compare these processes to those found in Faetar. One or both of these phenomena may be the source of the deletion process in Faetar. The possibility of multiple sources of linguistic change has not traditionally been considered: the family tree model allows for only one parent language for each daughter. However, newer analyses have posited multiple sources for some changes. One such example concerns the ancestry of the Tok Pisin focus marker yet, which may be considered to have a substrate source in the Tolai post-nominal focus marker [æt] or a superstrate source in English yet (Sankoff 1992). Guy suggested that both may be considered as input causing the change (Linguistics 660 lecture, 11/8/93).
7 Types of contact and types of change
Having provided a description of the contact situation, including the relative status of the languages involved, and descriptions of typologically distinct changes, the stage is set for a test of claims concerning the relationship between types of contact and types of change, made by linguists such as Thomason and Kaufman (1988), Van Coetsem (1988), and Guy (1990). Guy (1990) has reduced the types of language change to the following trichotomy: spontaneous (internal) change, borrowing (external, recipient language as agent), and imposition (external, source language as agent). Because many speakers are balanced Faetar/Italian bilinguals, it is difficult to state whether the changes which have taken place in Faetar are borrowings or impositions. Guy proposes social, psychological, and linguistic characteristics of change for each of these types. I will consider each of these to determine how best to characterize the events in Faetar. The case of balanced bilinguals is considered the limiting case for this type of trichotomy (Van Coetsem 1988, cited in Guy 1990:50), so the success of the model in such a case would provide further proof of its validity. Each of the above authors has made claims about types of change which will occur in different social settings, and these will be considered in light of the Faetar situation. Faetar has remained distinct from Italian, in spite of the high rate of bilingualism in the community and the economic advantages that Italian provides. Because of this unusual situation, the effects of the status of the language can be distinguished from that of the status of the speakers. Social contexts of the type present in Faeto (where the minority language, spoken by an isolated group of speakers, has high status within the speech community, although its speakers are not of higher status than others in the region) have not been discussed. Each of the above authors also suggests a "cline of borrowability," claiming that, in a given social context, certain linguistic elements are more likely to be borrowed than others (Weinreich 1953, Thomason and Kaufman 1988:50, Van Coetsem 1988:25-44, Guy 1990:58). I briefly discuss evidence of additional types of borrowing from Italian into Faetar. I show that Faetar has adopted many lexical items from Italian (Nagy 1993) and that the two languages have nearly identical phonemic inventories. I will also show that the syntax of the two languages remains distinct, focusing on differences in negation and subject pronoun usage. This stage seems to be in keeping with the proposed "borrowability" hierarchies which state that less stable items, such as vocabulary, will be borrowed more readily than the more stable items, such as syntactic structures. One major flaw in all of the claims about relative likelihood or frequency of borrowing at different levels of the grammar is that there is no metric for measuring the amount of borrowing. How can one directly compare the amount of phonological borrowing to the amount of lexical borrowing? Certainly, comparing the number of phonemes borrowed to the number of lexemes is not appropriate. I propose rather, that the ratio of borrowed items to "original" items can be compared for each level of the grammar. However, many questions remain to be answered. For example, should borrowings be ranked according to frequency of types or tokens? Does a borrowing such as lexically-marked length distinction from Italian into Faetar count as one change or must we count one change for each phoneme which now is distinctive with respect to length? Does the existence of initial and medial geminates count as one borrowing or two? How can we overcome the apparent contradiction that languages which, originally, have more in common with each other will necessarily be rated as having less influence on each other? These are the types of questions which must be answered before any of the "cline of borrowability" claims can be directly tested. My dissertation will make significant steps toward finding satisfactory answers. This type of classification system should prove useful in the analysis of historical changes which took place unobserved by today's linguists. Additionally, the contribution toward understanding the organization of our cognitive linguistic faculty which has already been begun by catalogers of types of language change can be greatly improved once quantitative data is available.
In this proposal, I have described how my investigation of linguistic change in a small speech community will be conducted. I combine historical linguistic and sociolinguistic methodology to describe the history of a relatively unstudied language in light of current variable patterns in the language. I use comparative reconstruction and dialect geography methodology to determine the ancestor of the language, then Labovian methods of analyzing synchronic variation to show how the language has changed since its inception, with respect to selected sound changes. This research serves a variety of purposes. In addition to providing a corpus of recorded speech for further research, new data on grammatical borrowing, and a test of recent theories of the interaction of contact situation type and linguistic change type, this project preserves a language which is destined to disappear shortly as more of its speakers leave the community. I hope that each of these achievements will aid in future linguistic research. Bibliography
Note: This bibliography contains references to all sources cited in the proposal, as well as others works that I have consulted or will consult in preparing this dissertation.
Abramson, Arthur. An analysis of initial geminates in Patanay Malay. Phonetic Science Congress. Actes du 5e Congrès internationale de Langue et litterature d'occitanie et d'études franco-provençales. 1964, 1967. published by G. Moignet & R. LaSalle. Amery, H. and S. Carthwright. 1987. First 100 Words. Tulsa: EDC. Anglade, Joseph. 1921. Grammaire de l'ancien provençale: phonétique et morphologie. Ascoli, G.I. "Schizzi franco-provenzali" in Archivio Glottologico Italiano, III (1878-1879). pp. 61-120. Bailey, Guy, Tom Wikle, Jan Tillery, and Lori Sand. (1991) "The apparent time construct." Language Variation and Change 3.3: 241-264. Barthe, Roger. 1970. Lexique Français-Occitan. Paris. Basboll, Hans. "Phonological Weight and Italian Raddoppiamento Fonosintattico." Rivista di Linguistica. 1989. v. 1.1. p. 5-31. Bertinetto, P.M. "La quantità vocalica e consonantica in italiano: verifica spettrografica." Actes du XIV Congrès Internationale de Linguistique et Philologie Romane. Naples 1974. Bourciez, Édouard. Éléments de linguistique romane. Paris: Klincksieck, 1956. Carosiello, John. "A Brief History of Faeto and Celle San Vito." ms. Castielli, Raffaele. Funtan' d' Fait' Testi integrali dei Cori con presentazione traduzione e note. Faeto 1975. Cortelazzo, Manlio (ed.), Profilo dei dialetto italiani . Consiglio Nazionale delle Ricerche, Centro di Studio per la Dialettologia Italian, Pisa:Pacini. 1974 -. DeRosa, Maurilio. Il Borgo Natio: Storia Diplomatica del Commune di Faeto in Terra di Capitanata. Molfetta 1934. De Salvio, A. "Relics of franco-provençal in Southern Italy" in Publications of the Modern Language Association of America XXIII (1918) pp. 45-79. Dourguin, Camille & Charles Mauron. 1987. Lou Prouvençau à l'Escolo Marseille: Lou Prouvençau à l'Escolo. Duraffour, Antonin. 1969. Glossaire des Patois Francoprovençaux. Paris: Editions du Centre National de la Recherche Scientifique. Durand, Bruno. 1973. Grammaire provençale. 4th edition. Association pédagogique "Lou Prouvençau a l'escolo." Fino, Carmela. 1970. "Il dialetto francoprovenzale di Celle S. Vito in provincia di Foggia." Master's Thesis: Università degli studi a Bari. Fragassi, Anna. 1977. Master's Thesis: Università degli studi a Bari. Gallucci, Concetta. "Provenzale o Franco Provenzale?" in Il Provenzale, August 1988. Gallucci, Pietro. Cenni di storia cronologica di Faeto. 1-Quaderni de Il Provenzale. Naples 1882. Gardette, Pierre. 1950. Altas linguistique et ethnographique du Jura et des Alpes du Nord. Paris: CNRS. Gardette, Pierre. Études de géographie linguistique. Strasbourg. 1983. Gilliéron, Jules & Edmond Edmont. 1902-1910. Atlas Linguistique de la France. Paris: E. Champion. Giustiniani, L. 1802. Dizionario geografico ragionato del Regno di Napoli, Napoli, (ristampa anastatica, ed. Bologna: Forni. 1969-1971), vol IV, S. V. "Faeto." Grandgent, C. 1927. From Latin to Italian: An Historical outline of the phonology and morphology of the Italian language. Harvard University Press. Guy, Gregory R. (1990) "The sociolinguistic types of language change." Diachronica. VII:1.47-67. Guy, Gregory R. (1991) "Explanation in Variable Phonology: An Exponential Model of Morphological Constraints," in Language Variation and Change 3:1. Hall, Robert A., jr., 1958. Bibliografia della linguistica italiana, Firenze: Sansoni. Hambert and Elugbe. "Acoustic correlates of fortis consonants." Han, Mieko. "The timing Control of Geminate and Single Stop Consonants in Japanese: A Challenge for Nonnative Speakers." Phonetica 1992; 49:102-127. Hermann, Eduard. "Lautveränderungen in der Individualsprache einer Mundart." Nachrichten der Gesellschaft der Wissenschaften zu Göttingen Phil.-His. Kl., 11:195-214. 1929. Hochberg, Judith. "/s/ deletion and pronoun usage in Puerto Rican Spanish." In D. Sankoff (ed.), Diachrony and Diversity. New York: Academic Press. pp. 199-210. Hock, Hans. 1991. Principles of Historical Linguistics. New York: Mouton de Gruyter. Holtus, Günter, Michele Metzeltin & Max Pfister. 1989. La Dialettologica Italiano oggi. Tübingen. Jaberg, Karl & Jakob Jud. (ed.) 1928-1940. Sprach- und Sachatlas Italiens und der südschweiz, Zofingen:Ringier. Jochnowitz, George. 1973. Dialect Boundaries and the Question of Franco-Provençal. Mouton: The Hague, Paris. Jud, Jakob. 1978. Romanische Sprachgeschicte und Sprachgeographie. Kattenbusch, Dieter. 1982. Das Frankoprovenzalische in Süditalien. Studien zur synchronischen und diachronischen Dialektologie. Tübingen: Gunter Narr Verlag. Kattenbusch, D. "Faeto und Celle: Frankoprovenzalische Sprachkolonie in Süditalien. Ein Beitrag zur soziolinguistik." in Italienische Sprachwissenschaft . Schwarze, (ed.) Saarbrücken 1979, Tübingen, 135-146. Kattenbusch, D. "Mort ou survie d'un dialecte?" Lengas: Revue de sociolinguistique 7-1980, 143-147. Kattenbusch, D. "Plurilinguismo nel sud d'Italia" Variation linguistique dans l'espace: Dialectologie et onomastique. Actes du XVIIème Congrès International de Linguistique et Philologie Romanes (Aix-en-Provence 1983) Vol. No 6, 1986. 399-412. Labov, William. "The social motivation of a sound change." Word 19:273-309, 1963. Labov, William. 1966. The Social Stratification of English in New York City. Washington, DC: Center for Applied Linguistics. Labov, William, P. Cohen, C. Robins and J. Lewis. 1968. A study of the non-standard English of Negro and Puerto Rican Speakers in New York City. Cooperative Research Report 3288. Vol I. Philadelphia: Linguistics Laboratory, University of Pennsylvania. pp. 123-157. Labov, William, Malcah Yaeger, and Richard Steiner. 1972. A Quantitative Study of Sound Change in Progress. Philadelphia: The U.S. Regional Survey. Lahiri, A. and J. Hankamer. "The timing of geminate consonants." Journal of phonetics 26, 327-338. 1988. Mandalari, M. "Una colonia provenzale nell'Italia meridionale, in G.B. Basile, Archivio di Litteratura Popolare, II, 1. (1884) Martin, Jean-Baptiste and Gaston Tuaillon. 1981. Atlas linguistique et ethnographique du Jura et des Alpes du Nord. Paris: CNRS. Melillo, Antonio. 1990. "La Tutela delle Minoranze Linguistiche." Master's Thesis from the Università degli studi di Bari. Melillo, M. (1981) "Briciole francoprovenzali nell'Italia meridionale." Vox Romanica 40: 39-47. Melillo, M. "Intorno alla probabili sedi originarie delle colonie franco-provenzali de Celle e Faeto" in Revue de Linguistique Romane. XXIII (1959) pp. 1-34 Melillo, M. "Il tesoro lessicale franco-provenzale odierno di Faeto e Celle in provincia di Foggia" in L'Italia dialettale, XXI (1956-7) pp. 49-128. Melillo, M. 1966. Lingua e società in Capitanata, Foggia. p. 69-78. Melillo, A. M. Storia e cultura dei francoprovenzali di Celle e Faeto: un Saggio Storico culturale di R. Castielli ; I Francoprovenzali di Capitanata ed Una Novella nel Francoprovenzale di Faeto di M. Melillo. Italy: Atlantica. 1978. Melillo, M. "Donde e quando vennero i franco-provenzali di Capitanata" in Lingua e storia in Puglia. Siponto: Centro Residenziale di Studi Pugliese. 1974. Meyer-Luebke. 1967. Grammatica storica della lingua italiana. Torino. Migliorino. Storia della lingua italiana. Minichelli, Vincenzo. 1994. Dizionario francoprovenzale Celle di San Vito e Faeto. Alessandria: Edizione dell'orso. Morosi, G. "Il dialetto franco-provenzale di Faeto e Celle, nell'Italia meridionale" in Archivio Glottologico Italiano, XII (1890-1892). pp. 33-75. Nagy, Naomi. (1992a) "A Geographic Analysis of the Origins of Faetar." The Penn Review of Linguistics 17:177-188. Nagy, Naomi. (1992b) "A Prosodic Description of Faetar." 1992. ms. Nagy, Naomi (1993). "Lexical change and language contact." Penn Review of Linguistics 17: 117-132. Philadelphia: University of Pennsylvania Department of Linguistics. Nagy, Naomi. (1994) "Language Contact and Change: Italian (?) Geminates in Faetar." to appear in Belgian Journal of Linguistics 9. Papanti, Giovanni "I parlari Italiani" Livorno, Vigo. 1875. Pellegrini, Giovan Battista. 1977. Carta dei Dialetti d'Italia, Pisa:Pacini. Pfister, Max. LEI. Lessico etimologico italiano, Wiesbaden, Reichert, 1979/ 1984. Piat, L. 1893 Dictionnaire Français-Occitanien, vol. I. Montpellier. Pulgram, Ernst. 1958. The Tongues of Italy: Prehistory and History. Greenwood Press, New York. Rebora, Piero. 1967. Cassell's Italian Dictionary. New York: Funk and Wagnalls. Recihstein, Ruth. "Study of social and geographic variation of linguistic behavior." Word 16:55, 1960. Regione Puglia. 1989. Faeto. Troia: Centro Regionale Distrettuale FG/31. Reynolds, Barbara. The Concise Cambridge Italian Dictionary. Cambridge University Press 1985. Reynolds, Bill and Naomi Nagy. (1994) "Phonological variation in Faetar: An Optimality Account." Chicago Linguistic Society 30-II: Papers from the Parasession on Variation and Linguistic Theory. Chicago. Ricerce sulla cultura locale Francoprovenzale per il XXX Concorso di Patois "Abate G.-B. Cerlogne." 1991-92. Ricerche sulla fauna nella cultura locale francoprovenzale per il XXIX Concorso di Patois "Abate G. - B. Cerlogne." 1990-91. Rickford, John and Faye McNair-Knox. (1992) "Addressee- and topic-influenced style shift: a quantitative sociolinguistic study." Perspectives on Register: Situating Register Variation within Sociolinguistics. ed. by Douglas Biber and Edward Finegan. Oxord: Oxford University Press. Riddle, Joseph E. 1849. English-Latin Lexicon. New York: Harper. Rohlfs, G. 1966. Grammatica storica della lingua italiana e dei suoi dialetti. G. Einaudi, ed. Rohlfs, G. "A proposito dei `franco-provenzali' in provincia di Foggia," in Studi di storia pugliese in onore di Giuseppe Chiarelli, Congedo, Galatina, 1973, II, 411-416. Rubino, Leonardo & Vincenzo Rubino. eds. Il Provenzale, Periodico della minoranza Franco Provenzale di Faeto e Celle S. Vito. Faeto, Italy. Salvi, S. "Le minoranze linguistiche in Italia," in U. Bernardi, Le mille culture, Roma, 1976, p. 138-150. Sankoff, Gillian. Linguistics 440 hand-out "Focus in Tok Pisin." 11/10/92. Savino, Luigi. Memorie storiche del comune di Celle S. Vito. Schmidt, Johannes. 1872. Die Verwandtschaftsverhältnisse der indogermanischen Sprachen. Wiemar: Böhlau. Schwan-Behrens. 1932. Grammaire de l'ancien français. Leipzig: O.R. Reisland. Sobrero, A. 1974. Dialetti Diversi, Lecce. Somerfelt, Alf. "Sur la propagation de changements phonétiques." Norsk Tidsskrift for Sprogvidenskap 4:76-128. 1930. Straka, Georges. 1972. Les Dialectes de France au moyen age et aujourd'hui; Domaines d'oïl et domaine franco-provençal. Paris: Klincksieck Thomason, Sarah and Terrence Kaufman. 1988. Language Contact, Creolization, and Genetic Linguistics. University of California Press. Traupman, John C. 1966. The New College Latin and English Dictionary. New York: Bantam. Valente, Vincenzo. 1975. Puglia. (volume 15 of M. Cortelazzo (ed.) Profilo dei dialetti italiani) Pisa: Pacini. Van Coetsem, Frans. 1988. Loan Phonology and the Two Transfer Types in Language Contact. Dordrecht: Foris. Zuccaro, L. 1894. Lucera et les colonies provençales de la Capitanata du siècle XIII. Foggia.
|
| This page was last modified by Naomi Nagy on 01/16/01 |