Allophones are defined by means of complementary distribution. As I understand it, a complementary distribution is a "mutually exclusive" relationship between two phones, with regard to a certain phonetic environment. That is, one of the phones will only be found in that environment (and nowhere else) while the other phone will never be found in that environment (but will be found somewhere else).

For example, In English, [pʰ] and [p] are allophones of the phoneme /p/, since [pʰ] can be found at the beginning of syllables ([pʰɪn]) and nowhere else. Likewise, [p] is never found at the beginning of syllables, but can be found in other positions ([spɪn]). Summarizing in a table:

Phoneme /p/ in English

Now, in Brazilian Portuguese, [w] and [l] are considered allophones of /l/. The reasoning is similar: [l] is never found at the end of syllables, but can be found in other places ([a.'kli.vi], "slope"). [w], on the other hand, is found at the end of syllables ([saw], "salt"). The problem is: [w] can also be found elsewhere ([a.'kwi;.fe.ru], "aquifer"). The table for this distribution, therefore, would be like this:

Phoneme /l/ in Brazilian Portuguese

This means that, in some environments both [w] and [l] can be found, as in the [a.'kli.vi] versus [a'.kwi.fe.ru] example. So, evidently, I'm failing to understand what "complementary distribution" really is. Any help to clarify the concept would be much appreciated!

  • 1
    It's not that simple as it might seem. Just because two phones are in complementary distribution, it does not automatically make them allophones of the same phoneme - compare English [h] and [ŋ]. Usually another criterion is necessary - of phonetic similarity - and even this one is not sufficient in certain cases.
    – Alex B.
    Jul 1, 2012 at 2:08
  • The thing is that there are some phonemes with limited (defective, restricted) distribution. And some languages resist a traditional account in phonological theory - compare Kabardian vowels.
    – Alex B.
    Jul 1, 2012 at 2:20
  • What Alex B. says certainly jibes with what I found at this link: sil.org/linguistics/GlossaryOfLinguisticTerms/… Jul 1, 2012 at 4:41
  • 1
    I always think about superman and clark kent. Where one is present, the 'other' necessarily cannot be even though they are 'the same person'. Does that help? Or think of two sides of a coin...
    – user1151
    Jul 1, 2012 at 15:29
  • Consider the words "mal" and "mau", the pronunciation is equal. But I could not find any examples of this in the middle of the word.. as there is "balde" but no "baude", "molde" but no "moude", it seems that there are no such words, and that something is different in those cases, when /l/ as [w] occurs in the middle of the word.
    – Tames
    Jul 1, 2012 at 18:14

2 Answers 2


I think you need to tease apart the concept of allophony from the concept of complementary distribution. As people have mentioned in the comments, the two don't necessarily go hand in hand--i.e. it is not necessarily the case that two phones that are in complementary distribution can be assumed to be allophones of the same phoneme (as Alex B. mentioned), nor is it the case that the distribution of two phones--each of which is a possible realization of a single phoneme--will never overlap in their distribution. It's the second assertion that is relevant for your Brazilian Portuguese example.

The "trick" is that sometimes different phonemes can have allophones "in common", i.e. Phoneme A might sometimes be realized as Phone Y, but Phoneme B might also sometimes be realized as Phone Y:

/A/ --> [X] in some environment, [Y] in some other environment

/B/ --> [Y] in some environment, [Z] in some other environment

This happens a lot in neutralizations and mergers. In languages with word-final devoicing, for example, maybe /d/ is realized as [t] word-finally but as [d] elsewhere. But /t/ would also be realized as [t] word-finally. If one encounters a [t] at the end of a word, one cannot be sure without additional information whether it is a realization of /t/ or /d/.

In your example, [w] is probably an allophone of some phoneme that can sometimes be realized as [w] and sometimes [l] (in some sense the "naming" of phonemes is arbitrary, but I'd probably call it /l/ since that seems to be the clear "elsewhere" realization) and [w] is probably also an allophone of a different phoneme (one that I would call /w/).

Now, it is true that allophones of the same phoneme must be in complementary distribution, but the two [w]s and the [l] in your table are not actually all realizations of the same phoneme. So, here is a revised assessment of their distribution, taking their underlying forms into account:

[l] (from /l/) not found at the end of a syllable, but found elsewhere

[w] (from /l/) found at the end of a syllable, but not found elsewhere

[w] (from /w/) found elsewhere (and perhaps also at the end of a syllable?)

So, the [l] and the first [w] are in fact in complementary distribution. I hope that makes sense!

  • 2
    It may also be the case that /w/ does not exist as a phoneme in BP, and that [a.'kwi;.fe.ru] contains a phoneme /kʷ/. I'm not saying this is the case - I don't know enough about BP to know. But it is a possibility.
    – Colin Fine
    Jul 2, 2012 at 10:37
  • 1
    @Colin you may be right. From the Wikipedia article on Portuguese phonology: "Bisol (2005:122) proposes that Portuguese possesses labio-velar stops /kʷ/ and /ɡʷ/ as additional phonemes rather than sequences of a velar stop and /w/" Jul 2, 2012 at 17:09

complementary distribution mean that the allophones of a particular phoneme occur in different phonetic environment , the phonetic environment determines which of the allophones is used in pronouncing a word . For example ,the two allophones of the phoneme /p/, namely , [pH] and [p-]are in complementary distribution the first occurs at the end of the word,as in Pat and Pot, the second may occur at the end of the ward , as in help. the fact of their complementary distribution means that these allophones do not occur in minimal pairs in English.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.