Comments on Test 1

Class Average 36.68; Range 13.5 - 51

The test yielded an very wide range of performance, ranging from full marks to well below 50%. An overview of the grading scheme and some comments on things to watch out for in future tests are given below.

Note that this test is worth 10% of the course grade, whereas the 5 homeworks completed so far are worth approximately 23% of the grade. Therefore, although the test may give a useful indication of how well you have mastered some of the main areas covered in the course, your homeworks are a greater contributor to your course grade.

Question 1: True/False

[5 points]

Take great care in answering these questions - although they are easy to answer, it is all too easy to answer them incorrectly.

Question 2: Multiple Choice

[5 points]

Question 3: Phonetic Transcription

[6 points: 0.5 deducted for each 3 errors]

This question was mostly done fairly well, but a number of common friends still occurred. Don't forget:

• To distinguish the high front tense vowel (as in "beat") from the high front lax vowel (as in "bit")
• To distinguish the voiced interdental fricative (as in "this) from the voiceless interdental fricative (as in "thing).

etc. ... see the web pages for the phonetics homework(s) and the tips page for test 1 for more examples of this.

Question 4: Phonetic Features

[4 points: 0.5 points per answer]

In specifying the features of a sound it is important to provide enough detail such that the features pick out only the sound that you are interested in. For example, the description "nasal" applies to three different stops in English, so information on place of articulation (e.g. labial, velar etc.) needs to be provided; similarly the description "voiceless alveolar" applies to both /t/ and /s/ in English, so information on whether the sound is a stop (/t/) or a fricative (/s/) is needed. This was the most common cause of problems.

Tip: when you describe a sound in terms of features, ask yourself whether there are multiple sounds that fit your description. If so, then you need to provide a more specific feature description.

Question 5: Phonological Analysis

[16 points: 2 points for each of 8 sections]

This was the question that really sorted people out, especially the first dataset. These problems were less complex than many of the practice problems that were discussed in class or review sessions, and the generalizations were similar to ones that we have seen before, so these problems should not have presented too many problems.

• /s/ --> [z] rule: affects the Spanish phoneme /s/ when it occurs immediately preceding a voiced consonant. Does not apply when either a voiceless consonant or a vowel follows. A small number of people also pointed out that the [z] allophone occurs exclusively in syllable-final position. This was strictly true for the dataset in the test, but it would be disconfirmed by a word such as "mas", which is probably pronounced [mas] in all dialects of Spanish.
• /d/ --> [th] rule: affects the Spanish phoneme /d/ when it occurs between two vowels.
• For each allophone, 1 point available for listing the environments in which the allophone occurs. Additional point available for stating an accurate generalization. No credit given for accurately stating that the sounds are in complementary distribution if no further justification for this conclusion is given anywhere in the answer (e.g. if no generalization is given in answer to parts a/b).
• Assimilation: a rule which voices a consonant when it immediately precedes another voiced consonant is clearly an example of an assimilation rule. The example of a rule which creates a fricative out of a stop when it occurs between two vowels is also a possible instance of an assimilation rule - a stop is pronounced as a continuant when it occurs between two continuants.

Note: if your generalization does not work for all of the forms in the dataset, then it is not the correct generalization, so you should look for another one.

Question 6: Morphophonology

[5 points]

Identifying the 4 forms of the Lithuanian prefix was easy. Extracting a maximally simple generalization caused more problems. The following simple statements can be made:

(i) When the stem begins with a stop, the prefix ends in a nasal with the same point of articulation.

(ii) When the stem begins with a fricative consonant, the prefix ends in the long vowel /a:/.

Pointing out both of these generalizations (or something similar) was needed in order to get full credit; a maximum of 3 points were available without these generalizations.

Question 7: The Language Instinct

[6 points: 3 points for each of two reasons]

Most people successfully identified two of the arguments that language is a human instinct, but in order to receive full credit an answer needed to be (i) clearly written, (ii) provide relevant supporting evidence.

You should always take care to read over what you have written and make sure that it is something that you expect to be understandable to somebody else and that it is a well-formed sentence (at least descriptively, if not prescriptively!).

Question 8: Genie

[4 points]

Most people gave a clear explanation of how Genie's difficulties in acquiring the grammar of English provide support for the existence of a critical period for language acquisition - this was enough for 3 points. However, the question also asked what we can learn from the aspects of language that Genie did learn relatively well, and only a small number people answered this part of the question. Associating labels with words, and creatively expressing complex meanings are both key features of human language which Genie seems to have done fairly well at learning, suggesting that there may not be a critical period for these aspects of language.

Question 9: Children's Spelling

[4 points: 2 points for clear explanation of each of two spelling errors]

The basic point, which almost everybody pointed out, is that children's spellings may more accurately represent the surface phonetics of English than standard adult spellings do. In order to receive full credit for this question, though, it was necessary to give specific reasons for the spellings used. Many of the child-adult spelling differences did not reflect situations where the adult spelling is arbitrary, but rather situations in which the adult spelling reflects the phonology rather than the phonetics of English. For example the use of a "d" in the spelling of words like "butter" and "hitter" reflects the American English flapping rule, which applies to /t/ when it occurs between vowels and following a stressed syllable. Also, the spelling of /sk/ sequences as "sg" reflects the effects of the aspiration rule in English, which leads to a phonetic difference between syllable-initial aspirated /k/ and unaspirated /k/ in other positions.