Phonemes

• are the smallest sequentially organised unit in the structure of a language,
• have the function of encoding the extremely large number - i.e. thousands - of morphemes (smallest meaningful units, e.g. roots, affixes, grammatical words) of a language with sequences constructed out of a small inventory - i.e. a couple of dozen - sounds.
• are characterised only in terms of distinctive features, i.e. a minimum of features required to distinguish the phonemes
• are realised concretely by allophones, which, unlike phonemes, are specified in full phonetic detail; examples:
  , i.e. as aspirated, or unaspirated, or nasally released plosive, depending on the context.
• is consequently sometimes defined as a bundle of distinctive features
• or as a set of allophones which are phonetically similar (like the p-allphones shown above) and in complementary distribution (i.e. different allophones do not occur in the same positions in the word).

As the smallest sequentially organised unit in the structure of a language, the phoneme is at the bottom of the following hierarchy of sequentially organised units:

Dialogue:

Dialogue: smallest interaction constituting sign

Text, utterance:

smallest dialogue constituting sign

Sentence:

sentence as smallest text constituting sign

Word:

smallest sentence constituting sign, with the following constituents:

1. inflected word (word-stem + inflection)
2. compound word (2 word-stems, optionally with interfix)
3. derived word (1 word-stem, with affixes (prefixes, suffixes)
4. morpheme: smallest meaningful sign

Phoneme:

smallest morpheme encoding sign.