I. Same sounds but different representations

Two or more languages might share the same sound or sounds but this does not mean that those languages organize these sounds in the same way.
 

[s] and [š]
Japanese	English
[šimasu] "do"	[slæš] "slash"

English:
 

• [s] and [š] can distinguish meaning:

          [mæš] "mash"
          [mæs] "mass"
 

• We cannot guess, by looking at the neighboring sounds, whether the word will start/end/contain an [s] or [š].

Japanese:

• Unlike English, these two sounds cannot distinguish meaning in Japanese:

• If you as a learner of Japanese make a mistake and say [simasu] instead of [šimasu], the meaning of the word would not change. You might, at most, be perceived as a foreigner and sound funny.

• However, in English if you say "I [s]aved my head this morning", you will confuse people.

• In Japanese, if there is [i] in a word, we will never see the sound [s] before it. We can also guess that [š] can precede it.

• [s] and [š] are contrastive and the occurrence of the two is unpredictable in English.

• In Japanese, on the other hand, we can predict the occurrence of one versus another. Thus, in Japanese, [s] and [š] are considered to be the "same" sound even though they may be phonetically distinct.

1. Of all the sounds in a language, which are predictable?
2. What is the phonetic context that allows us to predict the occurrence of these sounds?
3. Which sounds affect the meaning of words?

English: Consider the following words:

We all know that all of these words contain a /t/. But, do we really say the /t/s in the same way in all the three words?

Acoustic phonetic facts:

FACT #1:

• There is a burst or puff of air after the /t/ in "tip" that is absent in "stick".

• We learnt that this puff air is called aspiration and we represented it with a super-scripted [h].

• In English, aspiration occurs on all voiceless stops occurring as the first sound in a stressed syllable.

Examples:

[p^hat] vs. [spat] "pot" vs "spot"

[t^hek] vs. [stek] "take" vs. "stake"

Although, [t] in "tip" and [t] in "stick" are physically different, we, as speakers of English, consider both to be the same sound.

FACT#2
 

• The /t/ in "little" sounds a lot "softer" than the one in stick or tip, and voiced.
• In American English, this sound is actually pronounced as a flap (I will use [r] to refer to this flap sound since the original symbol doesn't appear on Netscape). But, we all know, at least at some psychological level, that this word contains a /t/ sound although we may not pronounce or hear it as such.

As speakers of English, we have the knowledge that /t/ is not always realized the same way but it may be pronounced several different ways. And, we overlook this difference when we learn words that contain this sound.

Hindi: Hindi also has aspirated and unaspirated stops:

    [p^hal] "knife edge"

    [pal] "take care of"

    [kapi] "copy"

    [kap^hi] "ample"

Remember that English also has aspirated and unaspirated stops and aspiration does not signal a
difference in meaning in English. Even if you pronounce "pill" as [pil] (as French speakers of English would be more likely to do), or "spill" as [sp^hil], the meaning of these words does not change.

However, aspiration is "contrastive" in Hindi. In other words, saying [pal] for "knife edge" instead of [p^hal] is like saying "shave" instead of "save".

So, while for English speakers, [p^h] and [p], or [t^h] and [t] are members of the same class, Hindi speakers cannot "overlook" these differences and distinguish meaning based on these difference.

A class of speech sounds that are identified by a native speaker as the same sound is called a phomeme.

The different phonetic realizations of a phoneme are called allophones.

Thus:

    [p^h] and [p] are allophones of the same phoneme in English.

    Whereas in Hindi, [p^h] and [p] are allophones of different phonemes.

Phonemes are the psychological (abstract) representations or units of actual physical realizations of phonetic segments.
 

II. Distribution of Speech Sounds

• If two sounds are separate phonemes, then they are contrastive (in terms of meaning).

• If the two phones are allophones of the same phoneme, then they are non-contrastive.

• To determine whether a given pair of sounds is contrastive, linguists look for minimal pairs.

Minimal Pair:

A minimal pair is a pair of words with different meanings with exactly the same pronunciation except for one sound that differs.

Examples:

[tek] vs. [tep] "take" vs "tape" (shows that [k] and [p] are contrastive in English)

[tim] vs [dim] "team" vs "deam" (shows that [t] and [d] are contrastive in English)

[kapi] vs [kap^hi] "copy" vs. "ample" (shows that [p] and [p^h] are contrastive in Hindi)
 

Examples:

Are [l] and [r] the same phoneme in English?

(Note: We are talking about the normal [r] in this case, not the flap sound for which I also used the symbol [r] above due to technical reasons.)

F Look for minimal pairs!

    [lif] "leaf"
    [rif] "reef"

    or

    [læk] "lack"
    [ræk] "rack".

Given that we have minimal pairs that contain [l] and [r], we can say that [l] and [r] are contrastive thus they are separate phonemes (i.e. they are NOT allophones of the same phoneme).

How about in Korean?

    [param] "wind"

    [irim] "name"

    [pal] "foot"

    [mal] "horse"
 
 

• In Korean, minimal pairs can never be found for [l] and [r] because these sounds can never occur in the same position in words.

• This set of data shows that [r] occurs between two vowels but [l] never occurs in the same environment.

• That is, [l] and [r] are in complementary distribution in Korean. They are mutually exclusive. In the same context, you can never find both.

-->  or  

We can say that Superman and Clark Kent are the "allophones" of the famous cartoon character.

Don't forget: When sounds are in complementary distribution, they are not contrastive. The replacement of one sound for the other does not change the meaning of the word.

Homework 4:

Exercises (1), (3) and (4) at the end of Chapter 7 in F&R