My personal experience suggests there is a background process taking place when interpreting a partially heard utterance. The sound is taken in, the conscious part of my mind begins the process of asking for a repetition because of imperfect reception and then a possible interpretation is delivered to my conscious mind for acceptance or rejection. Usually the logic can be seen, how a few words heard imperfectly can be combined to make sense, a partially heard name combined with a pronoun and a mental list of participants narrows the possibilities to one or two, but I don't have a sense of doing this analysis consciously. A second, almost a paraphrase, interpretation is offered up seemingly from the sub-conscious, "she's asking about Tony's birthday."

If you try to search for linguistics or language and error correction you will find out that there is a lot of concern about teacher behavior during language acquisition. That is not what I want to know about.

I would like to know if there is a subfield in cog-sci that deals with error correction specifically with regard to heard language. Somewhere I can begin a search. Sometimes all you need is the right search terms.

  • $\begingroup$ Please consult this answer on proper use of unconscious vs. subconscious in the current title of this question. $\endgroup$ – Artem Kaznatcheev Jan 31 '13 at 1:44
  • $\begingroup$ Sort of feel that my use of sub-conscious is entirely appropriate here since it reveals my status as a non-initiate. $\endgroup$ – timquinn Feb 1 '13 at 6:51
  • $\begingroup$ It is a positive use of ambiguity, no? $\endgroup$ – timquinn Feb 1 '13 at 6:51

A simple explanation for the phenomenon is top-down feedback. As the bottom-up acoustic/phonological input is coming in, there is top-down feedback based on your knowledge of the language, the situation, and all of the other contextual information, which is helping to constrain or inform your interpretation of the bottom-up signal. A classic example that is close to the example in your question is the phoneme restoration effect (originally demonstrated by Warren in 1970, and studied in depth by Samuel for decades): if a speech sound is replaced with an extraneous sound like a cough, listeners still report hearing the missing speech sound. Importantly, they report hearing exactly the sound that would belong in that context. So a manipulated word like "*eel"(where the * corresponds to the cough), sounds like "wheel" if it is the context "the *eel was on the axle", but the same word sounds like "peel" in the context of "the *eel was on the orange".

This kind of integration of multiple sources of information is typically called "interactive processing". For a review in the domain of speech perception see McClelland, Mirman, & Holt (2006).


McClelland, J. L., Mirman, D., & Holt, L. L. (2006). Are there interactive processes in speech perception? Trends in cognitive sciences, 10(8), 363–369. doi:10.1016/j.tics.2006.06.007

Samuel, A. G. (1981). Phonemic restoration: Insights from a new methodology. Journal of Experimental Psychology: General, 110(4), 474–494.

Samuel, A. G. (1996). Does lexical information influence the perceptual restoration of phonemes? Journal of Experimental Psychology: General, 125(1), 28–51.

Warren, R. M. (1970). Perceptual restoration of missing speech sounds. Science, 167(3917), 392–393.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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