More on Clausal Complementation and Selection Link
Project Directors: Mark Baker and Ken Safir , Rutgers University
This research initiative explores how predicates select for clausal complements. Natural languages most typically make available several different clause types which can function as the complements to a verbal (or adjectival) matrix predicate and it is frequently possible, in a given language, to predict which of the several clause types in the language the predicate will select based on the lexical semantics of that predicate. Native speakers appear to acquire the typology of clausal complementation in their language without explicit instruction, which raises familiar poverty of the stimulus questions that imply a great deal of tacit knowledge about the consequences of what a verb means for the syntactic form of the complement it selects. Although work on the semantic and syntactic selection by verbs that take nominal or prepositional complements has been a staple of linguistic work for many years (as, for example, in the work of Levin and Rappaport, 1995 and much work on thematic roles), relatively little research has sought to use crosslinguistic contrasts to identify the key factors, or even the key generalizations, that should inform our account of how speakers arrive at the classification of clausal complementation in their native language. Our proposal is to use the Afranaph resources to explore this question.
This is attractive in that the African languages permit us to investigate two types of crosslinguistic comparison: comparisons between languages that are broadly different, both in history and typology (e.g. across Afro-Asiatic, Niger-Congo, and Nilo-Saharan), and comparisons among languages that are broadly similar, but differ in the details of clausal selection in ways that reveal smaller cleavages in classification (e.g., across the Bantoid languages).
We focus on the following questions:
(A) How does the meaning of a verb influence the set of syntactically realized clausal complements it can select?
(B) How does the set of possible clausal complements influence the lexicalization of verb meanings? Do certain verbs mean what they do because of the complements they can select?
(C) Can the same complement clause have a different presuppositional commitment based on the sort of verb that selects it? In other words, do certain clause types mean what they do because they are selected by certain verbs?
(D) Is there a relationship between what a complement clause means and its internal syntactic structure? How is that relationship explained?
(E) How are the answers to questions (A-D) affected by the fact that the inventory of possible clausal complement in any given language can be strikingly different from the inventory of clausal complement types in other languages? Our leading hypothesis is that these relations are orderly.
The Systematic Selection Hypothesis: If the inventory of clausal complement types in a language is known, then the sorts of clausal complements a given predicate allows is predictable from its meaning.
It may turn out that the SSH is not defensible, or that it must be qualified in ways that will turn out to be systematic and interesting, but the strength of the SSH insures that only a thorough examination of the empirical patterns surrounding clausal complementation will allow us to decide the matter. If the SSH is supported in one form or another, it still remains to explain the results by answering questions A-E.
Insofar as our research must identify which differences in the meaning of a predicate influence its selection, the first order of business is to track the variation in clause types to see what the verbs that select them have in common. This in turn requires that we can identify what the clause types are in a given language, that is, to identify the inventory of possible distinctions in a given language based on the clause types it contains. Some languages have very few clause types while others provide a richer inventory. Thus our elicitations must be careful to explore not only the internal structure of clausal complements, but the fine-grained selectional distinctions that correlate with any detectable syntactic or semantic variation in the behavior of a complement clause.
Clausal Complementation Questionnaire
This research initiative explores how predicates select for clausal complements. Natural languages most typically make available several different clause types which can function as the complements to a verbal (or adjectival) matrix predicate and it is frequently possible, in a given language, to predict which of the several clause types in the language the predicate will select based on the lexical semantics of that predicate. Native speakers appear to acquire the typology of clausal complementation in their language without explicit instruction, which raises familiar poverty of the stimulus questions that imply a great deal of tacit knowledge about the consequences of what a verb means for the syntactic form of the complement it selects. Although work on the semantic and syntactic selection by verbs that take nominal or prepositional complements has been a staple of linguistic work for many years (as, for example, in the work of Levin and Rappaport, 1995 and much work on thematic roles), relatively little research has sought to use crosslinguistic contrasts to identify the key factors, or even the key generalizations, that should inform our account of how speakers arrive at the classification of clausal complementation in their native language. Our proposal is to use the Afranaph resources to explore this question.
This is attractive in that the African languages permit us to investigate two types of crosslinguistic comparison: comparisons between languages that are broadly different, both in history and typology (e.g. across Afro-Asiatic, Niger-Congo, and Nilo-Saharan), and comparisons among languages that are broadly similar, but differ in the details of clausal selection in ways that reveal smaller cleavages in classification (e.g., across the Bantoid languages).
We focus on the following questions:
(A) How does the meaning of a verb influence the set of syntactically realized clausal complements it can select?
(B) How does the set of possible clausal complements influence the lexicalization of verb meanings? Do certain verbs mean what they do because of the complements they can select?
(C) Can the same complement clause have a different presuppositional commitment based on the sort of verb that selects it? In other words, do certain clause types mean what they do because they are selected by certain verbs?
(D) Is there a relationship between what a complement clause means and its internal syntactic structure? How is that relationship explained?
(E) How are the answers to questions (A-D) affected by the fact that the inventory of possible clausal complement in any given language can be strikingly different from the inventory of clausal complement types in other languages? Our leading hypothesis is that these relations are orderly.
The Systematic Selection Hypothesis: If the inventory of clausal complement types in a language is known, then the sorts of clausal complements a given predicate allows is predictable from its meaning.
It may turn out that the SSH is not defensible, or that it must be qualified in ways that will turn out to be systematic and interesting, but the strength of the SSH insures that only a thorough examination of the empirical patterns surrounding clausal complementation will allow us to decide the matter. If the SSH is supported in one form or another, it still remains to explain the results by answering questions A-E.
Insofar as our research must identify which differences in the meaning of a predicate influence its selection, the first order of business is to track the variation in clause types to see what the verbs that select them have in common. This in turn requires that we can identify what the clause types are in a given language, that is, to identify the inventory of possible distinctions in a given language based on the clause types it contains. Some languages have very few clause types while others provide a richer inventory. Thus our elicitations must be careful to explore not only the internal structure of clausal complements, but the fine-grained selectional distinctions that correlate with any detectable syntactic or semantic variation in the behavior of a complement clause.
Clausal Complementation Questionnaire
Tense and Aspect in African Languages
Project Director: Silvester Ron Simango , Rhodes University , South Africa
The goal of this research project is to provide an in-depth investigation into the tense/aspect systems of African languages which will ultimately shed more light on the tense and aspectual properties of the languages and provide a basis for distinguishing between the two phenomena. From a purely descriptive point of view, TAM systems in African languages have not been systematically studied and more importantly, the temporal boundaries between the various tense forms have not been clearly established.
Africanists, and more specifically Bantuists have in the past pointed to the robustness tense systems found in African languages. Bamileke, for example, is said to have at least ten different tense forms (Anderson, 1983); iciBemba is also said have tenses that number into double figures (Chung and Timberlake, 1985); and ciCewa is said to have at least eight tenses (Mtenje, 1987). The categories that have been posited to characterize the tense systems in these languages include: ‘simple past’, ‘‘recent past’, ‘remote past’, ‘present’, ‘habitual’, ‘immediate future’ and ‘remote future’. However, the temporal boundaries between, say, the different types of ‘past’ have not been clearly defined in the literature.
A relatively recent study by Simango (2003) seems to suggest that there may well be far fewer tense types in these languages than has previously been thought and that certain tense forms that have been posited can be accounted for in terms of the aspectual properties of the language in question. This line of inquiry has revealed that ciCewa also makes the same kind of past tense divisions as ciNsenga and that iciBemba distinguishes between hodiernal, hesternal, and pre-hesternal past tense forms. Other languages such as Sotho and Swati seem to have only one form of past but have two ways of expressing pastness on the basis of factors considered here.
Of course additional factors may also be at play in determining what forms are appropriate for conveying information that relates time of event to time of utterance, and the proposed study would hopefully reveal what these are. Using some version of the Afranaph questionnaire or some other appropriate tools the study seeks to reveal how different African languages cluster temporal relations by the use of tense marking and to document how different tense forms in different languages convey different meanings in different languages.
Those interested in serving as native-speaker linguist consultants for this project should begin by consulting the Tense and Aspect Questionnaire (TAQ), which can be viewed and downloaded below or on the Become a consultant page. Those who decide to participate should then follow the instructions on the Become a consultant page, where details concerning how to participate (and remuneration) are provided. If you do decide to become a consultant, be sure to send a consent form which can be downloaded on that page.
TAQ [pdf] [doc]
The following languages have TA data entered in the database so far:
References
Anderson, Stephen C. 1983. Tone and morpheme rules in Bamileke-Ngyemboon. Doctoral dissertation, University of California, Los Angeles.
Chung, Sandra, & Alan Timberlake. 1985. Tense, aspect and mood. In Shopen, T. (ed.) Language Typology and Syntactic Description vol III: Grammatical categories and the lexicon. Cambridge:
Cambridge University Press.
Comrie, Bernard. 1985. Tense. Cambridge: Cambridge University Press.
Miti, Lazarus. 2001. A linguistic analysis of ciNsenga. Cape Town: CASAS.
Simango, Silvester Ron. 2003. Reanalysing past tense categories in Bantu.
Malilime: Malawian Journal of Linguistics 3: 67-84.
The goal of this research project is to provide an in-depth investigation into the tense/aspect systems of African languages which will ultimately shed more light on the tense and aspectual properties of the languages and provide a basis for distinguishing between the two phenomena. From a purely descriptive point of view, TAM systems in African languages have not been systematically studied and more importantly, the temporal boundaries between the various tense forms have not been clearly established.
Africanists, and more specifically Bantuists have in the past pointed to the robustness tense systems found in African languages. Bamileke, for example, is said to have at least ten different tense forms (Anderson, 1983); iciBemba is also said have tenses that number into double figures (Chung and Timberlake, 1985); and ciCewa is said to have at least eight tenses (Mtenje, 1987). The categories that have been posited to characterize the tense systems in these languages include: ‘simple past’, ‘‘recent past’, ‘remote past’, ‘present’, ‘habitual’, ‘immediate future’ and ‘remote future’. However, the temporal boundaries between, say, the different types of ‘past’ have not been clearly defined in the literature.
A relatively recent study by Simango (2003) seems to suggest that there may well be far fewer tense types in these languages than has previously been thought and that certain tense forms that have been posited can be accounted for in terms of the aspectual properties of the language in question. This line of inquiry has revealed that ciCewa also makes the same kind of past tense divisions as ciNsenga and that iciBemba distinguishes between hodiernal, hesternal, and pre-hesternal past tense forms. Other languages such as Sotho and Swati seem to have only one form of past but have two ways of expressing pastness on the basis of factors considered here.
Of course additional factors may also be at play in determining what forms are appropriate for conveying information that relates time of event to time of utterance, and the proposed study would hopefully reveal what these are. Using some version of the Afranaph questionnaire or some other appropriate tools the study seeks to reveal how different African languages cluster temporal relations by the use of tense marking and to document how different tense forms in different languages convey different meanings in different languages.
Those interested in serving as native-speaker linguist consultants for this project should begin by consulting the Tense and Aspect Questionnaire (TAQ), which can be viewed and downloaded below or on the Become a consultant page. Those who decide to participate should then follow the instructions on the Become a consultant page, where details concerning how to participate (and remuneration) are provided. If you do decide to become a consultant, be sure to send a consent form which can be downloaded on that page.
TAQ [pdf] [doc]
The following languages have TA data entered in the database so far:
References
Anderson, Stephen C. 1983. Tone and morpheme rules in Bamileke-Ngyemboon. Doctoral dissertation, University of California, Los Angeles.
Chung, Sandra, & Alan Timberlake. 1985. Tense, aspect and mood. In Shopen, T. (ed.) Language Typology and Syntactic Description vol III: Grammatical categories and the lexicon. Cambridge:
Cambridge University Press.
Comrie, Bernard. 1985. Tense. Cambridge: Cambridge University Press.
Miti, Lazarus. 2001. A linguistic analysis of ciNsenga. Cape Town: CASAS.
Simango, Silvester Ron. 2003. Reanalysing past tense categories in Bantu.
Malilime: Malawian Journal of Linguistics 3: 67-84.
The Morphosyntax of Bantu Nouns
Project Director: Tarald Taraldsen
This Afranaph Sister Project is an outgrowth of an investigation of the class prefixes on Bantu nouns using the formal relations between different kinds of agreement markers and subparts of the nominal prefixes as a window onto the underlying morphosyntactic structure of the class prefixes on nouns. This led to analyses formulated within a nanosyntactic framework which associate the morphemes occurring in nominal prefixes with complex syntactic structures. This work also yielded some insight into the construction of pronouns and demonstratives in Bantu.
Currently, our research efforts are in part directed towards improving on those initial analyses both on the conceptual and the empirical side. As part of this, we will need to focus more on getting data from Bantu languages outside the Nguni languages that provided our initial data base. We also want to deepen our understanding of how class prefixes and agreement markers interact with their syntactic environment. On the one hand, this concerns the alternation between different “allomorphs” of the agreement markers, e.g. the relative distribution of the class 1 subject agreement markers a and u, which seems to be conditioned by the clause type the verb finds itself in. On the other hand, we want to see to what extent the presence/absence of the augment, realized as the initial vowel in Nguni, correlates with interpretive and syntactic properties, and to what extent the alternation between augmentless and complete forms have any counterpart in Bantu languages that lack initial vowels. Therefore, we will shift our empirical focus away from paradigms and increasingly concentrate on the syntactic and semantic properties of sentences.
Since our research will now lead us to investigate things like the difference between clauses in the participial mood, the subjunctive mood and the principal mood, the structure of relative clauses, the structure of negation and scope properties, to mention just a few, we expect that our research interests will coincide enough with those of other potential Afranaph users that our empirical results will be useful also to them.
Those interested in serving as native-speaker linguist consultants for this project should begin by consulting the Noun Class Prefix Questionnaire (NCPQ), which can be viewed and downloaded below or on the Become a consultant page. Those who decide to participate should then follow the instructions on the Become a consultant page, where details concerning how to participate (and remuneration) are provided. If you do decide to become a consultant, be sure to send a consent form which can be downloaded on that page.
NCPQ [pdf] [doc]
This ASP is located at the Center for Advanced Study in Theoretical Linguistics at the University of Tromsø . It is funded by Tromsø Research Foundation and the University of Tromsø, and run by Tarald Taraldsen in collaboration with Michal Starke .
This Afranaph Sister Project is an outgrowth of an investigation of the class prefixes on Bantu nouns using the formal relations between different kinds of agreement markers and subparts of the nominal prefixes as a window onto the underlying morphosyntactic structure of the class prefixes on nouns. This led to analyses formulated within a nanosyntactic framework which associate the morphemes occurring in nominal prefixes with complex syntactic structures. This work also yielded some insight into the construction of pronouns and demonstratives in Bantu.
Currently, our research efforts are in part directed towards improving on those initial analyses both on the conceptual and the empirical side. As part of this, we will need to focus more on getting data from Bantu languages outside the Nguni languages that provided our initial data base. We also want to deepen our understanding of how class prefixes and agreement markers interact with their syntactic environment. On the one hand, this concerns the alternation between different “allomorphs” of the agreement markers, e.g. the relative distribution of the class 1 subject agreement markers a and u, which seems to be conditioned by the clause type the verb finds itself in. On the other hand, we want to see to what extent the presence/absence of the augment, realized as the initial vowel in Nguni, correlates with interpretive and syntactic properties, and to what extent the alternation between augmentless and complete forms have any counterpart in Bantu languages that lack initial vowels. Therefore, we will shift our empirical focus away from paradigms and increasingly concentrate on the syntactic and semantic properties of sentences.
Since our research will now lead us to investigate things like the difference between clauses in the participial mood, the subjunctive mood and the principal mood, the structure of relative clauses, the structure of negation and scope properties, to mention just a few, we expect that our research interests will coincide enough with those of other potential Afranaph users that our empirical results will be useful also to them.
Those interested in serving as native-speaker linguist consultants for this project should begin by consulting the Noun Class Prefix Questionnaire (NCPQ), which can be viewed and downloaded below or on the Become a consultant page. Those who decide to participate should then follow the instructions on the Become a consultant page, where details concerning how to participate (and remuneration) are provided. If you do decide to become a consultant, be sure to send a consent form which can be downloaded on that page.
NCPQ [pdf] [doc]
This ASP is located at the Center for Advanced Study in Theoretical Linguistics at the University of Tromsø . It is funded by Tromsø Research Foundation and the University of Tromsø, and run by Tarald Taraldsen in collaboration with Michal Starke .
DP Positions in African Languages
Project Directors: Vicki Carstens, University of Missouri, Michael Diercks , Pomona College, Loyiso Mletshe , University of the Western Cape, Juvénal Ndayiragije, University of Toronto, Justine Sikuku, Moi University
A central concern of syntactic theory has long been to explain and predict the distribution of nominal expressions (DPs), and their involvement in morpho-syntactic relations. Where can they occur? When can they move, control agreement, and bear Case? The study of Indo-European (IE) languages has yielded strong generalizations upon which the theory is based, based on empirical contexts such as subject-verb agreement, properties of infinitives, raising constructions and passive constructions. Carstens (2011) and Diercks (in press) show that these generalizations do not hold of Bantu languages. Some additional research points towards the same conclusion for various non-Bantu languages of Africa (see for example Ura 1998; Baker and Willie 2010; and Ouali, 2007). A large-scale cross-linguistic study is therefore well-motivated from a theoretical standpoint and has the potential to drive serious reconsideration of our theories of DP licensing as a component of the universal human faculty of language.
Given the broad range of empirical contexts in which DP licensing must be investigated, the data that arise out of this project should be of interest to anybody whose work intersects with those areas; people interested in complementation patterns will likely find our data on raising and infinitives versus tensed clauses relevant, while people working on inversion constructions will be interested in our data on locative inversion, subject-object reversal, and impersonal passives. At first we expect our investigations to focus on Bantu languages, but we are very interested in expanding beyond this family to non-Bantu Niger-Congo and Afro-Asiatic.
Those interested in serving as native-speaker linguist consultants for this project should begin by consulting the DP Positions Questionnaire (DPPQ), which can be viewed and downloaded below or on the Become a consultant page. Those who decide to participate should then follow the instructions on the Become a consultant page, where details concerning how to participate (and remuneration) are provided. If you do decide to become a consultant, be sure to send a consent form which can be downloaded on that page.
You can access the DPPQ here: [pdf] [doc]
Navigating our site
The languages for which we have collected data are arranged on separate pages in this site called our Case Files. The case file for each language in our study has the data elicited by our standard questionnaires, supplemented by follow-up elicitations designed to explore areas that are richer in the language in question than the same area may be in some other language. In addition, many case files have other features, such as a short grammar sketch of the language, a bibliography, or short summaries or sketches of the empirical results achieved for the language in question, as well as posted publications of works of interest on the language of interest (e.g., the Kinande Dictionary) a sketch of the anaphora system (highlighting aspects of anaphora in the language that appear to raise interesting theoretical or analytic issues). For further information about what is in completed case files and how best to use them, see About the Case Files or go directly to our Case Files page to view the case files we have so far.
Another major feature of our site is the online Afranaph Database which permits sophisticated search and manipulation of the data we have collected. More information can be found in About the Database and a tutorial is available in How to use the Database.
Those expect that those who intend to use any information on our site should read our Fair Use and Citation page for appropriate guidelines.
Interested users can access our Anaphora Questionnaire (AQ) to see our basic field elicitation document. Information about our Elicitation of Primary Data, about how to Contact Us, about how to Become a Consultant, and our Plans for the Future of the site, as well as Our History and Prospects are all available with a click on the topics listed on the lefthand side of your screen.
Another major feature of our site is the online Afranaph Database which permits sophisticated search and manipulation of the data we have collected. More information can be found in About the Database and a tutorial is available in How to use the Database.
Those expect that those who intend to use any information on our site should read our Fair Use and Citation page for appropriate guidelines.
Interested users can access our Anaphora Questionnaire (AQ) to see our basic field elicitation document. Information about our Elicitation of Primary Data, about how to Contact Us, about how to Become a Consultant, and our Plans for the Future of the site, as well as Our History and Prospects are all available with a click on the topics listed on the lefthand side of your screen.
