This book presents a new theory of grammatical categories - the Universal Spine Hypothesis - and reinforces generative notions of Universal Grammar while accommodating insights from linguistic typology.
SUMMARY This collection’s purpose is to bring together two subfields, morphology and language contact, to provide into cohesive analyses of language systems or situations of language contact an area where there’s been relatively limited work to date (e.g. Boretzky & Igla 1994, Gardani 2008, Wilkins 1996). Each of the 14 articles in this collection contributes to this intersection of morphology and language contact and come from differing perspectives focusing on how situations of language contact lead to morphological change or rearrangement and in particular, contact influences on bound morphology. The book itself is organized into six parts headed by either geographical area or language family groupings: Amerindia, Austronesia, Balkan (and beyond), Romance, Slavic (outside the Slavic core area), and Africa. Preceding these articles is a preface by the editors (Martine Vanhove, Thomas Stolz, Hitomi Otsuka, & Aina Urdze) detailing the goals of the project coming out of the international conference on Morphologies in Contact at the University of Bremen in October 2009. The volume closes with a list of contributors and separate indexes for authors, languages, and subjects.
Part I Amerindia contains three articles. Marianne Mithun (“Morphologies in contact: form, meaning, and use in the grammar of reference”), Françoise Rose (“Borrowing of a Cariban number marker into three Tupi-Guarani languages”), and Claudine Chamoreau (“Spanish diminutive markers –it-/-ita in Mesoamerican languages: A challenge for acceptance of gender distinction”).
Mithun argues that lexical forms do not always copy (‘get borrowed’) from one language to the other, but instead a morpho-semantic pattern can spread through language contact. She looks at northern California languages, primarily Yuki and Wappo which are related, and compares them to Pomo (primarily) and also Wintun languages, also unrelated. The strongest evidence is a grammatical marking pattern on pronouns. Both Yuki and Pomo have an agent-patient argument pronominal system (semantic-based), while Wappo, instead, has a nominative-accusative system. The actual forms of the pronouns vary between Yuki and Pomo, but the grammatical marking patterns mirror each other. Mithun argues that Yuki innovated its case marking system (including pronoun selection) based on contact with Poma, while Wappo lacked these innovations. The forms themselves did not get copied. Other evidence of these types of innovations is found with the inclusive/exclusive first person plural distinction, third person referential (only) type distinction, and reference across sentences.
Rose discusses the bound plural morpheme (*komo) in three Tupi-Guarani languages: Wayampi, Emerillon, and Zoᶦé. This morpheme, which was used as a collective marker, came into three languages through contact with the Cariban language family. No other bound morphemes have been borrowed into these three languages (and no other Tupi-Guarani language uses this morpheme). Rose discusses whether this bound morpheme comes from ancient contact between the Cariban and Tupi or more recent contact, and dismisses historical evidence for ancient contact. The Cariban languages use this form across the entire family as a collective marker suggesting that more recent contact is more likely the source of the innovation. Wayampi, Emerillon, and Zoᶦé are discussed regarding their structural similarities with the Cariban languages.
Chamoreau discusses diminutive markers (-ito/-ita) in Spanish which encode both gender and diminutive status. Some Mesoamerican languages have borrowed this marker although they already have diminutives. The author argues for a structural motivation for this borrowing, that is, that this category (diminutive) is conveyed via a suffix. Gender distinction is not carried over to the recipient language if it had no gender to begin with -- Only Yucatec Maya has native grammatical gender distinctions. Mesoamerican languages vary in how they natively form diminutives. Tepehua uses a phonological process of sounds being articulated further forward in the mouth, Tepehua and Purepecha use a separate word, and Nahuatl uses a suffix. The author categorizes the marker’s borrowing into three patterns: in Tepehua, it is found only on names (sometimes restricted to Spanish names) or Spanish lexical borrowings, in Purepecha, Central Mexicano, and Mexicanero de la Sierra Madre Occidental, only one marker (--ito) is borrowed meaning diminutive (so no gender distinction) but (--ita) is still found in lexical borrowings, in Yucatec both markers are borrowed to encode gender. Diminutive doubling is also found throughout these languages.
Part II Austronesia has only one article, Thomas Stolz (“Survival in a niche: On gender-copy in Chamorro (and sundry languages)”). The author argues that Chamorro, spoken in the Commonwealth of the Northern Marianas Islands and Guam, has adapted (at least sex-based) gender including gender agreement with adjectives from Spanish. While gender is not frequently transferred via language contact, Stolz argues for that case with Chamorro. While Chamorro is no longer in contact with Spanish, post-contact Chamorro was changed immensely by earlier contact with Spanish. Gender markers are only found on Spanish loans and Spanish adjectives and other modifiers agree with masculine gender unless the head is semantically feminine. Gender may only be analyzable on a small subset of the lexicon, but the author provides evidence of productive gender agreement in Chamorro and not just memorized Spanish loan forms.
Part III Balkan (and beyond) presents four articles: Evangelia Adamou (“Verb morphologies in contact: evidence from the Balkan area”), Birgit Igla & Irene Sechidou (“Romani in contact with Bulgarian and Greek: replication in verbal morphology”), Angela Ralli (“Morphology in language contact: verbal loanblend formation in Asia Minor Greek (Aivaliot)”), and Lars Johanson (“Mood meets mood: Turkic versus Indo-European”).
Adamou provides a typology of verb morphology under contact for Romani (mainly), but also Greek, Turkish, and Pomak. Language contact typologies include borrowing, grammatical replication, loan verb marker, and innovation. Form and meaning are transmitted during borrowing. Meaning and structure but not form are transmitted during grammatical replication. Loan verb markers occur when a form changes its meaning. Innovation is the process of lexicalization for a different meaning. Paradigm transfer of Turkish bound person morphemes into Romani is a case of borrowing. Backward diffusion occurs in Turkish and Slavic when borrowed morphemes apply to native recipient language verbs (Romani). A case of novel creation involves an obligatory past tense evidential marker in Turkish being adopted in Romani meaning ‘allegedly’. Loan verb markers from Turkish verbs are borrowed into Romani, but the evidential function is lost. A Greek verb marker loses its aorist function in Pomak. Forward diffusion involves borrowing loan verb markers from one language and affixing to verbs from a third language. The Greek aorist marker applies to English verbs in Angloromani although it has lost the aorist function. Nashta (Slavic) becomes more like Greek in function and structure (but not form) in some verbal inflections.
Igla and Sechidou suggest that Romani has replicated an analytic reflexive structure from Bulgarian and a synthetic passive structure from Greek. Bulgarian and Greek use the same marker in both the passive and reflexive structures. Greek has both analytic and synthetic reflexives, although the synthetic reflexive is more dominant pattern, while Bulgarian has only analytic patterns. The split pattern in Romani is argued to be made possible due to Romani speakers in Greece and Bulgaria. Greek Romani speakers replicate the synthetic reflexive construction used by Greek speakers. This allows two structures, both an analytic and a synthetic reflexive structure, to be used by Greek Romani speakers. Likewise, Bulgarian Romani speakers extend a reflexive pronoun to synthetic (Class II) verbs which already have a bound suffix for intransitives. Gradually the reflexive pronoun became associated with the intransitive/passive suffix, thereby ensuring an analytic structure for Romani passives.
Ralli discusses Turkish verb loans and loanblends in Aivaliot. Language-internal considerations (Greek structure) trump external factors presented by Turkish loans. The degree of bilingualism does not influence the morphological integration of such verbs and verbs are borrowed with the same frequency of nouns. A variety of Greek verbalizer morphemes appear on Turkish loan nouns while only one such morpheme is associated with Turkish loan verbs. Phonological reasons are suggested as being the reason that this morpheme is associated for Turkish origin verbs and not nouns. A Turkish past tense morpheme -di is always found on the Turkish origin verbs, but is left unanalyzed by Aivaliot speakers, providing a phonological environment in which only one Greek verbalizer can be affixed. While Ralli suggests that Greek-internal structural factors are most important, she concedes that external factors from source languages also do matter. Turkish loans in Greek can take Greek affixation without a problem, unlike English loan verbs. English loan verbs are found in a light verb (‘do’ + verb) construction. This structural difference is not attributed to bilingualism, because while bilingualism is not prevalent among Turkish and Greek speakers, it is prevalent among Greek-American speakers (those who employ the light verb structure).
Johanson suggests that mood markers and their structures in Turkic languages resemble those of Indo-European languages. Some convergence of these language families is evident, because these structures are similar enough due to frequent copying. Synthetic mood markers such as optative, voluntative, hypothetical, necessitative, and purpose attach as suffixes to a lexical verb in Turkic. These structures are represented by periphrastic modal constructions in which the lexical verb has a mood marker attached. If the subject is the same in both clauses, then an uninflected verbal noun is found, while the verbal noun is person marked in constructions which have different subjects. The “non-canonical” pattern, however, includes a junction between the two clauses. Examples from Ottoman Turkish, Cypriot Turkish, Azeri, Iraq Oghuz, Khorasan Oghuz, Gagauz, and Colloquial Turkish are given as evidence in showing these patterns.
Part IV Romance contains three articles: Michele Loporcaro (“Contact-induced change in personal pronouns: some Romance examples”), Immacolata Pinto (“The influence of loanwords on Sardinian word formation”), and Mauro Tosco (“Swinging back the pendulum: French morphology and de-Italianization in Piedmontese”).
Loporcaro discusses the role of contact in personal pronoun change in Italo-Romance languages. He compares case marking on personal pronouns in Standard Italian (and other Romance languages) with Latin. Some case distinctions are neutralized (contact-induced). Objects are derived from Latin datives in Logudorese Sardinian, but Standard Italian takes the accusative forms. Campidanese uses nominative forms for all functions due to contact with Catalan. The neutralization of some of these case distinctions is argued to be due to contact, because this would be less likely to occur outside the contact situations (Thomason 2003). Gender neutralization is contact induced in Logudorese and Gallurese. ''Contact put the gender marking system under pressure, in the first place, and this yielded a predictable result (a convergent system in the replica language, copying that of the model language) for all other word classes (p. 222).''
Pinto analyzes Sardinian morphological processes. Sardinian has very few productive prefixes (three), and gains no new ones through loanword adaptation. External factors play an important role in this diffusion. Pinto quantifies occurrences of prefixes and suffixes (among other morphological processes) of native and non-native words in a corpus. These are categorized into three “layers”: inherited Latin, complex Sardinian (with affixation), loan layer (mainly from Spanish and Catalan). Pinto discusses the percentage of occurrences of various affixes (prefix, suffix, etc.) in each particular layer. No prefixes are gained via loanword adaptation, although suffixes are. Socio-historical reasons are claimed to be most important here, but genetic motivations are also argued to play a role. Pinto claims that prefixes were prevalent in Latin and other Romance languages due to written texts.
Tosco discusses the “ausbauization” of Piedmontese, a process of language change due to revitalization and planning. These changes have been influenced by the acceptance of neighboring French and a rejection of Italian influence. Previously, Piedmontese existed in a stable multi-glossic situation with both Italian and French. With the rise of the Italian national state, Piedmontese has recently been endangered of becoming too “Italianized” and French has become a ''foreign'' language. Tosco suggests Piedmontese has undergone a process of ausbauization as evidenced by orthographic choices, variety choice, and corpus planning. Lexical form/meaning correspondences of words illustrate this process. Socio-historical factors matter most: ''Ultimately, it is ideology and politics which make the difference, not linguistics” (p. 261).
Part V Slavic (outside the Slavic core area) consists of two articles: Antonietta Marra (“Contact phenomena in the Slavic of Molise: some remarks about nouns and prepositional phrases”) and Lenka Zajícová (“Language contact, language decay and morphological change: evidence from the speech of Czech immigrants in Paraguay”).
Marra analyzes a Slavic variety, Molise, suggesting that contact with Romance helped establish preposition use in Molise. Gender marking of nouns is discussed, along with preposition use and case-marking. The neuter gender in Molise disappears and previously neuter forms become either masculine or feminine. Loanwords from Romance keep their original gender when they are integrated into Molise. Native Molise had a complex case system of inflectional morphology, but lacked prepositions, while Romance lacked the case-marking system, but provided prepositions. Modern Molise integrates both prepositions and case-marking. Interference from an 'in' prepositional phrase allows accusative marking on the noun, but if a Romance numeral loanword precedes the noun, then case is not marked. A reorganization of the case system has taken place, especially in regards to locatives. Both stative and motion functions of locatives are neutralized, marked by accusatives.
Zajícová discusses a Czech community in Paraguay whose Czech is heavily influenced by contact with local Spanish. This bilingual community is composed of first, second, and third generation speakers. Contact of Czech and Spanish drives speaker-creativity leading to morphological innovation. Grammatical systemization is not fully established, but there are some trends in speaker idiolects. Degrees of bilingualism vary: fluent speakers of Czech, rusty speakers, semi-speakers, and remembers (not analyzed). Three forces of language change are suggested processes: analogy, other language influence, and attrition. Fluent speakers of both Czech and Spanish show morphological changes. They use singular noun forms instead of plurals and adopt a masculine suffix. Spanish roots are used in Czech sentences. Spanish bound morphemes are found on Czech roots in Czech discourse. Spanish possessive pronouns co-occur with Czech ones. Semi-speakers lack ‘systematization’, that is, Czech use is highly creative and idiolectal. Attrition does not eliminate case morphology in Czech, but instead rearranges morphology in various patterns based on idiolect. Certain forms of specific words are lost, not the case ending itself.
Part VI Africa has only one article, Martine Vanhove (“Roots and patterns in Beja (Cushitic): the issue of language contact with Arabic”). Vanhove discusses a root and pattern typology in Beja, which has a morphological system similar to Arabic. Other Cushitic languages lack this similarity with Arabic and convergence between Beja and Arabic is argued to be the result, rather than borrowing, copying, or replication from Arabic forms. Bilingualism of men (there is less bilingualism among women) in both Beja and Arabic plays a role in convergence. “Beja is the only Cushitic language that uses qualitative ablaut in the stem as a derivational device, for the expression of various semantic and voice derivations (p. 315).” Beja differentiates between active and reflexive voices by alternating vowels, while other Cushitic languages do not. Beja and Arabic have root and consonant systems showing structural convergence, although the forms and meanings differ. This morphological system is less developed in Afar and Saho (other Cushitic languages), so the influence of Arabic on Beja is substantial.
EVALUATION “Morphologies in Contact” successful sheds light on a rarely researched topic. It contains numerous perspectives, methodologies, and case studies, but each investigation focuses on language contact and functional aspects of morphology.
A central issue addresses in each article whether internal factors, such as structural considerations of the languages that copy or replicate morphemes under contact, are more important or if instead external factors matter more, such as bilingualism, length of contact, and socio-historical influence. While each article addresses the nuances of both internal and external factors involved in a contact situation, certain authors are more firmly on one side (or the other) in this debate.
Chomoreau and Ralli stress the importance of language internal structure. Chamoreau emphasizes that the adaptation of a Spanish diminutive marker did not permit the adaptation of grammatical gender in languages which did not already have a gender system, and suggested that the diminutive marker fulfilled a structural purpose in these languages and not as a result of degree or length of contact. The cross-linguistic evidence is well organized and utilized. Ralli makes a relevant point regarding degree of bilingualism and structural adaption. Bilingualism is not prevalent among Turkish and Greek speakers although Greek inflection is found on Turkish verbs, but prevalent among Greek-American speakers who cannot inflect English verbs with Greek affixation. This is given as evidence to indicate that structural reasons trump sociolinguistic factors of bilingualism, which is particularly satisfying.
On the other side of this debate is presented in articles by Tosco, Pinto, and Zajícová. External factors such as socio-historical settings and bilingualism are the central reasons guiding morphological influence from a contact situation. Tosco pays close attention to ausbauization, which is a very active form of language planning and revitalization. The current socio-historical situation of Piedmontese involves a active rejection of Italian influence and acceptance of French forms into the variety. Pinto suggests that Latin or French prefixes were diffused into modern varieties of Sardinian via texts, although this, I believe, is dissatisfying in terms of being a main motivating factor or at least hard to support. Zajícová tells the beginning of a story of potential language shift in a bilingual community. The morphological implications based on the bilingual fluency of speakers in such a community is particularly interesting, although if language shift completes where does that leave the status of Czech bound morphology on Spanish lexical items or vice-versa?
Another relevant topic gets at how morphology is influenced via language contact. Some authors discuss situations where the phonological forms themselves do not copy. Mithun, Johanson, and Vanhove discuss replication and convergence of morphological patterns caused by bilingualism. Other authors focus on a specific morpheme that is copied during a contact situation and its distributions, functions, and structures. Rose, Chamoreau, Stolz, Loporcaro, Marra, and Igla & Sechidou focus entirely on one copied morphological process/ morpheme. These instances of morpheme copying are easier to argue than replication and convergence, although Mithun provides a clear discussion of historical reconstruction that is particularly compelling. Adamou proposes typologies of different types of contact induced morphological changes in the Balkans. One author, Zajícová, even tackles individual morphological considerations in a bilingual community, which provides a viewpoint that focuses on the language faculty itself.
This volume is particularly appealing for those looking for empirical studies on morphology and language contact and it provides a direction for future research in that area.
REFERENCES Boretzky, Norbert & Igla, Birgit. 1994. Interferenz und Sprachwandel. In Benedikt Jessing (ed.), Sprachdynamik. Auf dem Weg zu einer Typologie sprachlichen Wandels. Vol. III: Interferenzlinguistik, 7-139. Bochum: Brockmeyer.
Gardani, Francesco. 2008. Borrowing of inflectional morphemes in language contact. Frankfurt: Peter Lang.
Thomason, Sarah. 2003. Contact as a source of language change. In Brian Joseph & Richard Janda (eds.), The handbook of historical linguistics, 687-712. Malden, MA & Oxford: Blackwell.
Wilkins, David. 1996. Morphology. In Hans Goebl et al (eds.), Kontaktlinguistik: Ein internationales Handbuch zeitgenössischer Forschung, Vol. 1, 109-117. Berlin & New York: Walter de Gruyter.
ABOUT THE REVIEWER:
Christopher Batteen is a PhD Candidate in Linguistics at the University of Minnesota. His research interests are language contact and Bantu linguistics. His dissertation work is on code-switching in Chichewa/English online discussions approaching it from a generative approach. He seeks to finish this work in 2013 and then find meaningful employment in research and teaching.