----------------------------------------------------------------- Comments: Somehow the encodings were different between the two files you sent --- I could see the extended ascii characters in the test suite, but not in the write up. Hausa has non-verbal sentence structures which aren’t covered by the SVO pattern. These sentences use prepositions, a stabilizer, or an existential ‘there is/are’ word to indicate state. Also, Did you leave off the sentence after "Also"? I don’t think Hausa contains auxiliaries. Tense, aspect and mood are represented in a pre-verbal complex that the reference grammar refers to as the PAC (person-aspect complex). It is made up of a person pronoun connected to a tense/aspect/mood marker. The author notes that it corresponds to INFL. I’m having trouble deciding what to call the PAC. Does this count as a particle? An inflectional marker? Since the PAC must precede the verb, I added ungrammatical test cases where it is in different positions. I think there are two ways to treat the PAC --- one is as an auxiliary and the other is as an affix that the orthography just happens to separate from the verb (some Bantu languages are apparently written this way, in so-called "disjunctive" orthographies). The way to tell the difference would be if anything can ever intervene between the PAC and the verb. Or rather, if any open class thing can ever intervene there, since closed class things (e.g., negative markers) might also just be disjunctively written affixes. If you go for the affix analysis, you'll want to assume your morphophonological "preprocessor" nixes the extraneous space between the PAC and the verb, or perhaps replaces it with a -. It's good to include the non-verbal sentences in your test set, but you might not get to them this quarter. I'm assuming that two unmarked NPs next to each other is not a . Not a what? Determiners can occur before or after the head of a NP. Demonstratives like that/those/this precede the noun. A noun can have determiners preceding and following it, i.e. ‘wancàn gidâ-n’, that house-the, ‘that very house.’ I think this means that "the" isn't a determiner, but rather an affix indicating definiteness. You might want to analyze the demonstratives as determiners, or perhaps just as demonstrative adjectives... Getting the grammar to handle the ungrammaticality of examples like this one: Source: author Vetted: f Judgement: u Phenomena: {Pronouns, Argument Optionality} ita takàn dafà àbinci. ita takàn daf-à àbinci. she 3fsg.HAB cook food 'She cooks food' Is going to be interesting. I think we'll probably just have to invent a feature, call it PRO and have verbs insist on [PRO -] subjects. As mentioned before, a simple personal pronoun (with no modifiers) must be dropped if it is the subject. This implies a transformational analisys. A surface oriented statement would be more like "Simple personal pronouns (with no modifiers) cannot appear as a subject. Subjectless sentences must be used instead." What kind of modifiers "rescue" such sentences? Objects can't be dropped, as they are not in agreement with the verb. It's not actually true that agreement is a prerequisite for argument drop (crosslinguistically). Is the statement above based on something from your reference grammar, or is it a conclusion that you draw? (It seems to be incompatible with the next observations you make). I found it difficult to construct ungrammatical examples for sentential negation, since putting the negative markers in a different place just negates a different part of the sentence. How about examples with just one ba (at the begining or the end) or sentences with ba and the stabilizer in the wrong order? Since the pronouns change form for object, subject and genitive positions, does that mean that Hausa has a case system? I know in English the case system doesn’t change the form of nouns, but it does change pronouns in accusative case. Yes, that sounds like a case system at about the English level of complexity. I wouldn't bother with the genitives for this grammar, but the subject v. object distinction is worth capturing as case. (But, how do you know what the pronouns look like in subject position, if they're always deleted. What do the examples that have them overt look like?) I’m very confused about how verbs are formed in Hausa. There are 8 verb grades, and I don’t have a clear understanding of what they indicate. The grade tells something about the semantics of the verb, but it also seems to be just a convenient way of grouping together similar verb forms (or ending vowels). I think I need to read the section on verbs again, but do you know any resources that might help? Not specifically, but I would suggest talking with Kelly, since she's working on another Bantu language. You might also do a search online or in the MLA bibliography on "Hausa verb" to see what kind of scholarly works you can turn up. Also, you can bring the reference grammar by my office and we can look at it together :-) Source: author Vetted: s Judgement: u Phenomena: {Word Order} mùtûm ginà yaa gidaa. mùtûm gin-à yaa gidaa. man built 3msg.COMP house 'the man built a house.' "Vetted: s" is supposed to mean that the example was taken from some published source. Examples that you construct are "Vetted: f" until you run them past a native speaker, at which point you can list the judgments associated with those examples are "Vetted: t" Source: author Vetted: f Judgement: u Phenomena: {Word Order} yarò yaa kooyàa manà lìssaafìi yarò yaa kooy-àa manà lìssaafìi boy 3msg.COMP teach us.obj math 'the boy taught us math' Source: author Vetted: f Judgement: u Phenomena: {Word Order} yarò manà yaa kooyàa lìssaafìi yarò manà yaa kooy-àa lìssaafìi boy 1pl.obj 3msg.COMP teach math 'the boy taught us math' Source: author Vetted: f Judgement: u Phenomena: {Word Order} yarò lìssaafìi manà yaa kooyàa yarò lìssaafìi manà yaa kooy-àa boy math 1pl.obj 3msg.COMP teach 'the boy taught us math' Isn't one of the above grammatical? Formatting problem here: Source: author Vetted: f Judgement: {Determiners} tsoohuwaa wani taa zoo. tsoohuwaa wani taa zoo. old-woman some 3fsg.COMP came 'Some old woman came.'