Homework 1

How would a linguist probably respond to each of the following statements? A few sentences should suffice for each one.

1. Many educated Americans sometimes forget that the word data is plural and say, for example, her data is useless. This usage is incorrect.
   For most speakers of American English, data is not plural; it is singular (or neutral with respect to number).
2. Many languages have a standard form and in addition a number of dialects.
   The standard is also a dialect.
3. Because writers have more time to think about the language they're producing than speakers, it makes more sense to look at written than spoken language if we want evidence for the grammar of a language.
   Spoken language is more basic, but written language is not merely a visual representatio of spoken language; it has its own interesting properties.
4. Because it is not possible to divide signs into smaller units, analogous to the primitive segments of spoken language, signed languages are not true human languages.
5. We can discover many of the grammar rules of a language by asking native speakers what they are.
   Native speakers are rarely conscious of the rules.
6. Children are able to figure out what words mean because they get enough information from speakers and from the context.
   Many linguists believe that children don't get enough information from the context and that they need innate help in the form of constraints on what words can mean in order to learn word meanings. (Not everyone agrees though.)
7. Prescriptive grammar, such as the "correct" usage of who and whom, is described in terms of rules; the "natural" grammar spoken by children when they begin school is not rule-governed.
   "Natural" grammar is also rule-governed (though the rules are not conscious). (Note two uses of grammar, one referring to the knowledge a native speaker has, the other to a formal characterization of a set of rules. The latter is either descriptive or prescriptive. But it would not make sense to say that a speaker has a prescriptive, or even a descriptive, grammar.)
8. Most of what adult American English speakers know about their language they learned in school.
   Actually children may learn a lot of language in school. Most of this probably comes from the input they receive from reading: some grammar and a lot of vocabulary. But the basic system of the language (what most linguists care about) is already in place when they start school.
9. European languages tend to be more abstract and difficult to investigate than African languages.
10. When it is difficult to analyze a particular language, it is often useful to use the grammar of another language as a standard.
11. Children vary considerably in how well they know their language because their parents vary in how seriously they teach it to them.
    Most linguists believe that children learn language using a combination of the input they receive and the constraints that come to them from Universal Grammar. Because of the help they get from UG, differences in the quality of the input don't make much difference. In any case it doesn't appear that first languages can be taught. (But UG and an innate language faculty are just one view; not everyone agrees with this perspective. Also competence, the unconscious knowledge of the grammar of a particular language, is not innate. UG is not a grammar but constraints on the forms that grammar may take.)
12. Except for words that imitate sounds, language never directly reflects the nature of the world in any obvious way.
    This question concerns the relationship between form and meaning, whether this relationship is iconic or arbitrary. Languages may be iconic in a number of ways other than onomatopoeia.
13. Because speakers often restart their sentences in midstream (for example, she really liked the book she lent me ... I lent her), a complete grammar of the language must include rules for restarts.
    Because restarts (which are of course perfectly normal) reflect performance, not competence, most linguists are not interested in them. (Some psycholinguists are, however, and they have discovered that there are interesting constraints on how restarts happen.)
14. Languages differ in every possible way; only genetically related languages can be expected to be similar to one another.
15. English is the total of the linguistic behaviors of all speakers in the English linguistic community.
    English is what the different speakers share (possibly with some provision for major differences).
16. Arabic and Swahili have many similar words. This is strong evidence that the two languages are related.
    No, they could also be similar because of contact. (But note that a typological relationship would not lead to similarity in vocabulary.)