TIPS ON SOLVING PHONEMIC ANALYSIS PROBLEMS


Your task is to determine whether [b] and [p] are allophones of the same phoneme or different phonemes. 

1. Look for a minimal pair which differs on the presence of [p] vs. [b]. You only need one such minimal pair. This is sufficient to tell you that [b] and [p] are different phonemes. Your work is done.

2. Assuming there are no minimal pairs, start on stating the environments in which each is found. You can use the following notation or any other one that works for you.

        [b]                 [p]


         #_i                #_o

         a_i                i_i

        s_o                s_#
          
         e_#                a_l

"#" = word boundary


3. Once you have your phonetic contexts (or environments) listed, study them. Can you make any generalizations about the classes of sounds which come before and/or after the sounds in question?

4. If the two sounds appear to occur in the same environments, their distributions are overlapping. Unless you have evidence for free variation*, you should conclude that these sounds are different phonemes.

*What's evidence for free variation: what looks like a minimal pair- two words which differ phonetically on only one sound, but whose meaning remains the same.

5. If you can find a conditioning environment, that is, an environment in which one sound is found and the other is not, than you can conclude that the two sounds are in complementary distribution and they are thus allophones of the same phoneme. 

6. The final step is to select a basic allophone and a derived allophone. The basic allophone will be the one which has the broader distribution, i.e. is found in the most environments. The derived allophone is the one which is most restricted in the phonetic contexts in which it appears.  Once you have identified which is the basic (or underlying) sound, and which the derived one, you can state a rule which captures the conditioning environment.

An example of a rule: /p/ becomes [p^h] in stressed-syllable initial position.

Another example: /s/ becomes [z] before a voiced sound.

Note: be careful when you state the conditioning environment. Try to capture it with the most general class as possible (e.g. high vowels, rather than [i] and [u]), making sure, however, that there are no other data in your set which contradict that statement.