Words that are more frequent tend to be short, and conversely, words that are less frequent tend to be longer, containing more segments. This pattern makes language more efficient by balancing speaker effort versus the amount of information provided to the listener. But not all segments are equally informative: listeners identify words from the speech stream incrementally, continually updating their lexical search as the phonetic signal unfolds. As a consequence, segments earlier in words tend to contribute more disambiguating information to lexical access than later segments, depending on what similar lexical competitors exist. It is then expected that languages should not only optimize the total number of segments in words, but also how informative those segments are in disambiguating from other words in the lexicon. Here the speaker will show data from a range of languages that this is the case: words that are on average less frequent have relatively more informative early segments, while preserving a 'long tail' of more redundant later segments. 

The speaker will review the recent work suggesting that this asymmetry in segment information distribution across the word may influence the evolution of phonological rules which impact lexical identification. In a typologically-balanced sample of 50 languages, one can find that phonological rules which neutralize lexical distinctions (e.g., word-final obstruent devoicing in German) are common at word-ends, but very rare at word-beginnings. 

Both of these patterns are consistent with the hypothesis that language change is shaped by a tendency for speakers to reduce redundant phonetic material while preserving more informative material.