Phonemes vs. Allophones: Complementary Distribution

1. The Mental vs. The Physical: Phonemes and Allophones

In the last lesson, we focused on phonetics, the physical reality of speech sounds. Now, we move into phonology, the study of how those sounds are systematically organized in a language.

The central idea is that of the phoneme: an abstract mental representation of a sound that serves to distinguish meaning. We denote phonemes with slashes, like /t/.

When we actually speak, we produce allophones, which are the concrete, phonetic realizations of a phoneme. We denote allophones with square brackets, like [t] or [tʰ]. A single phoneme can have one or more allophones.

This video provides an excellent introduction to these concepts.

As the video explains, allophones of a single phoneme are like Clark Kent and Superman. They are different manifestations of the same entity, and they appear in different, predictable environments. You find Clark Kent in an office and Superman at the scene of a disaster; you never find them in the same place at the same time. This is the core of complementary distribution.

2. Finding the Pattern: Two Types of Distribution

To determine whether two sounds are different phonemes or allophones of the same phoneme, we analyze their distribution—the set of phonetic environments in which they can occur.

2.1 Contrastive Distribution → Different Phonemes

Two sounds are in contrastive distribution if they can occur in the exact same phonetic environment and changing from one to the other creates a word with a different meaning. This is the primary test for identifying distinct phonemes.

The tool for this test is the minimal pair: two words that are identical in every way except for one sound in the same position.

• Example: pat [pæt] vs. bat [bæt]
  • The environment [__æt] is identical.
  • Swapping [p] for [b] changes the meaning.
  • Conclusion: /p/ and /b/ are distinct phonemes in English.

The following clip from the video you started watching explains this test.

2.2 Complementary Distribution → Allophones of the Same Phoneme

Two sounds are in complementary distribution if they are phonetically similar and never occur in the same phonetic environment. Their "distributions" are complementary, like two puzzle pieces that fit together but don't overlap.

• Example: English /t/ has several allophones.
  • [tʰ] (aspirated, with a puff of air): occurs at the beginning of a stressed syllable. Ex: top [tʰɒp]
  • [t] (unaspirated): occurs after an [s]. Ex: stop [stɒp]
  • [ɾ] (flap): occurs between two vowels when the second is unstressed. Ex: butter [bʌɾɚ]

You cannot swap these sounds and keep the same pronunciation environment. For instance, you can't have [ɾ] at the beginning of a word ([ɾɒp]) or [tʰ] after an [s] ([stʰɒp]). Their environments are mutually exclusive. Swapping them doesn't change the word's meaning, it just sounds unnatural.

This concept can be visualized clearly. The set of all possible phonetic environments is partitioned, with each allophone occupying its own distinct subset.

The allophones of the English phoneme /p/ provide another classic example.

3. A Systematic Approach to Phonemic Analysis

Given your background, you can think of this analysis as an algorithm. We follow a set of procedures to arrive at a conclusion about the relationship between two sounds.

The following reading lays out this procedure very clearly.

Now, let's see this procedure in action with a dataset from a language you haven't seen before.

As you saw in the Georgian example, the process is:

1. List environments: Systematically record the sounds immediately preceding and following each target sound.
2. Check for minimal pairs: If any identical environments exist, the sounds are in contrastive distribution and are separate phonemes.
3. Find the pattern: If the environments are mutually exclusive, the sounds are in complementary distribution. Identify the phonetic feature that defines the environments (e.g., "before front vowels" vs. "elsewhere").
4. Conclude: Based on phonetic similarity and complementary distribution, the sounds are allophones of the same phoneme.

4. Stating the Rule

The final step in the analysis is to formalize the distribution as a phonological rule. This rule makes the relationship explicit. The standard format is:

/Phoneme/ → [Allophone] / Environment

This is read as: "The phoneme is realized as the allophone in this specific environment." The environment is shown with an underline representing the position of the sound.

• For the Georgian example, the rule for [l] would be:
  /ɫ/ → [l] / __ [front vowel]
• The default allophone (the one that appears in the most varied or "elsewhere" environments) is usually chosen to represent the phoneme. In this case, [ɫ] is the default, so the phoneme is written as /ɫ/.

The following video provides another example from Persian and shows how a final distribution statement is formulated.

5. Practice: Analyzing Canadian French

Let's apply this procedure to a problem. Given your interest in French, this dataset from Canadian French will be relevant.

Problem: Consider the sounds [t] and [ts] (written as [c] in the resource) in the following data. Are they allophones of the same phoneme or do they represent two different phonemes? If they are allophones of the same phoneme, state the distribution.

Data from e647f, Exercise 6. Note: [ts] is a voiceless alveolar affricate.

1. [tu] 'all'
2. [abuti] 'ended'
3. [teleɡʁam] 'telegram'
4. [tʁɛ] 'very'
5. [tɛl] 'such'
6. [tɛ̃bʁ] 'stamp'
7. [kyltsyʁ] 'culture'
8. [minyt] 'minute'
9. [tsimid] 'timid'
10. [tsy] 'you'
11. [tsit] 'title'
12. [tsyb] 'tube'

Follow the procedure: list the environments for [t] and [ts], look for patterns, and draw a conclusion.

Conclusion

Today we've made a fundamental distinction between the physical reality of a sound and its function within a language system.

Key Takeaways:

• Phonemes (/p/) are abstract, meaning-distinguishing sound categories.
• Allophones ([p], [pʰ]) are the concrete phonetic realizations of a phoneme.
• Contrastive distribution, identified by minimal pairs, proves that two sounds are separate phonemes.
• Complementary distribution, where two sounds appear in mutually exclusive environments, indicates they are allophones of the same phoneme.
• Phonemic analysis is a systematic process of listing environments to discover the distribution and formulate a rule.

You are now equipped to look at raw phonetic data and determine the underlying phonological system. This is a core skill for a historical linguist.

In our next lesson, "Analyze phonemic data to determine if two sounds are allophones of the same phoneme or distinct phonemes," we will continue to hone this skill by working through more complex data sets, solidifying your ability to apply these principles to new linguistic puzzles.