American Scientist

To the Editors:

I just finished reading Lise Abrams's intriguing article "Tip-of-the-Tongue States Yield Language Insights" (May-June). Despite lecturing in biology for some 35 years, I had never pondered the neural complexity of uttering a sentence.

Being an avid music lover, this question came to mind: Is there a musical analogue of tip-of-the-tongue (TOT) state for words? Suppose you asked someone "How does that tune go in Porgy and Bess where someone sings about not taking the Bible literally (It Ain't Necessarily So)?" If the respondent replies that the tune is on the tip of her tongue (or larynx?), are there "lists" of other tunes one could hum, some of which would help to resolve the "TOT" and others which would hinder the resolution?

Jules Lerner
Chicago, IL

To the Editors:

I read with much interest and enjoyment Lise Abram's article on the tip-of-the-tongue phenomenon. I was particularly taken by figure 5, comparing groups of different ages, in particular the green bars representing TOT resolution with phonologically unrelated words as retrieval cues. It is tempting to speculate on the relatively greater retrieval in the oldest group. If the phonologically unrelated words are not serving as cues, then perhaps some unspecified process is occurring to a greater degree among the oldest subjects in the interval between original and repeated general knowledge questions. What would happen if there were a fourth condition in the study in figure 5, in which a completely nonlinguistic task was interposed? For example, if subjects had to work on a visuospatial task that would preclude devotion of resources to language, would the same degree of TOT resolution occur?

William R. Lovallo
VA Medical Center and University of Oklahoma Health Sciences Center
Oklahoma City, OK

Dr. Abrams responds:

Dr. Lerner poses a fascinating question that has never been empirically studied, to my knowledge. It certainly seems plausible that we could have tip of the tongue (TOT) states for music, as we have them for other stimuli besides verbal material, such as odors, and there is a "tip of the fingers" phenomenon in deaf signers. The difficulty of studying priming from similar tunes comes in quantifying relatedness, as tunes can be related to one another on a number of dimensions, such as pitch, tempo and contour. I am a piano player myself, and the relationship between language, memory and music has always interested me. A great book about music and the brain is Musicophilia by Oliver Sacks.

Dr. Lovallo also raises a good point. The results shown in figure 5 are the first where that age difference emerged as significant (although there was a trend in that same direction in my other study with two groups of older adults). I think that a number of explanations are possible. One is that which Dr. Lovallo suggested, where the act of having a linguistic task is especially beneficial to the oldest adults. I'm not sure why this would be; possibly it has to do with older adults' increased vocabularies and networks of semantic knowledge. However, given that the linguistic tasks I have used are relatively simple (such as reading words aloud and rating words' pronunciation difficulty), this explanation isn't convincing to me.

No one has used a nonlinguistic task, so it would be interesting to see its effects on TOT resolution. Given the implicit nature of TOT resolution, in that we spontaneously come up with the missing word without direct awareness of the cause, my hunch is that engaging in any form of cognitive processing "gets the juices flowing," similar to insight that occurs suddenly when solving a difficult problem. Therefore, the nature of the intervening material as linguistic or nonlinguistic may be irrelevant to TOT resolution, but it's an empirical question.

It would be important to ensure that younger and older adults are equivalently capable of switching between linguistic tasks (such as answering general knowledge questions, having TOT states) and nonlinguistic tasks (such as a visuospatial task) in a short period of time.

Another explanation for the finding in figure 5 could be that the oldest adults are able to use mediated priming in this case and can get phonologically related information indirectly from the unrelated words. Again, I don't know why they would be able to do so more than younger adults.

A more trivial explanation is that the age difference in vocabulary (older adults have vocabularies that are superior to those of younger adults) may cause younger adults to more often respond with "TOT" on targets that they don't really know (to avoid continually saying "don't know"), thus they don't resolve them spontaneously because they never knew the words in the first place. This would make their baseline TOT resolution (following unrelated words) lower than in older adults.