My world has been shattered! This week I learned of two new studies that diverge from the past decade's prevailing evidence that our brains process speech and music in almost indistinguishable ways. (I wonder if we've been experiencing some degree of confirmation bias. (That was just an excuse to link to a great recent article by Jonah Lehrer. (This is just an excuse to embed a tertiary pair of parentheses!)))
The bad news is that I don't have access to the full text of these articles, so I won't be able to analyze the science as much as the simplified conclusions. The good news is that this post won't be as long as it otherwise would've been. The other good news is that other bloggers have discussed these articles in enough detail (and the abstracts themselves are well enough written) that I can still talk about them.
Victoria Williamson (in the blogroll) introduces us to research by Rogalsky et al. in her post, Music and Language – not as much overlap as we thought? These authors used fMRI to monitor the brains of subjects as they listened to meaningless yet grammatically sound sentences, meaningless scrambled sentences, and basic novel melodies. Comparing the meaningless sentence processing to the melody processing revealed that "sentences elicited more ventrolateral (lower and outer) activation, whereas the melodies elicited a more dorsomedial (upper and inner) pattern" in the temporal lobe. Yes, there was also overlap* in the auditory cortex (a region of the temporal lobe), but that is attributable to the primary processing of sound.
*Additional data even demonstrated that within that overlap, patterns of speech and music processing were still distinct, yet I strongly suspect this is due to the differences in acoustical properties between the melodic sounds played and the spoken voice.
The final information gained from the neurological processing of scrambled sentences furthered their support for the argument that "basic hierarchical processing for music and speech recruits distinct cortical networks." Vicky words it well:
The authors conclude that the distinguishable patterns of activity likely reflect the different auditory features present in speech and music: Also, the fact that while language is largely dependent on grammar, music is more dependent on pitch and rhythmic contours. Finally, the 'end game' of language and music are fundamentally different; the former is to derive combinatorial semantic representations and the latter is to drive acoustic recognition and perhaps emotional modulation.(Sorry, that's a slightly modified quote because I think she made a typo.)
Alas, I have a criticism (one which may have been addressed in the full text). What if a basic melody isn't the best correlate to a sentence? My suspicion is that the syntax of a chord progression is much more analogous to the syntax of a sentence. Perhaps the differences found in hierarchical processing would be less pronounced, or different, if harmony were compared instead of melody.
Another set of researchers (also in California) seem to have been simultaneously trying to answer one of the questions that Rogalsky's research begs. Parker Tichko (also in the blogroll), this time, frames the work best:
Previous research proposed that whether an auditory stimulus is perceived as music or speech is assumed to be a result of its acoustic properties. Thus many studies focused on decoding, or reverse-engineering, the acoustic components of speech and music. However, recent research has proposed that music and speech might be processed differently on a neural level, suggesting that it is not necessarily the acoustic properties of auditory stimulus that result in the perception of music or speech, but rather independent neural correlates.In this, Deutsch et al.'s, research, the authors played a trick on their subjects. I'll gloss over the details, but they essentially played a recording of a sentence and then repeated a part of it ten times (sometimes in a modified set, sometimes unmodified) and asked their subjects to rate whether the sounds were more like speech or song.
It's amazing how much the voice sounds like it is singing by the end of that, isn't it? It came as no surprise to read that the authors' survey revealed agreement among their subjects. Interestingly, they sometimes had a subject repeat the phrase after listening to it ten times, and they usually sung it back. Strikingly, that singing was closer in precision to a musically notated approximation of the original. This suggests that as this voice's speech began to sound like song, the importance of pitch and rhythm became so strong that the recording eventually sounded more musically imperfect then the subject's own rendition!
So, who wants to see the results of fMRI while a subject is listening to such a speech-to-song illusion? I DO.
Corianne Rogalsky, Feng Rong, Kourosh Saberi, and Gregory Hickok (2011). Functional Anatomy of Language and Music Perception: Temporal and Structural Factors Investigated Using Functional Magnetic Resonance Imaging The Journal Of Neuroscience : 10.1523/JNEUROSCI.4515-10.2011
Diana Deutsch, Trevor Henthorn, and Rachael Lapidis (2011). Illusory Transformation From Speech To Song Journal of the Acoustical Society of America, 129 (4), 2245-2252: 10.1121/1.3562174