> I assume you are relying on Russell & Nilsen, PNAS 94 (1999), 2660-2664 to demonstrate that the traveling wave is a > fact.
No, I wasn't. I was referring to the same paper that I referred to on
the cochlear list: PNAS 97, 11751-11758. But they reported the 65 dB postmortem
threshold in earlier papers as well. By the way, the paper you reference
is from 1997.
> That is, the measurements simply show that there is an accumulating
phase delay. This delay could be due to a traveling
> wave, but there are other possibilities. It does not disprove my conjecture.
I did not base the proof of the TW on phase accumulation, but on tuning curve shape. These tuning curve shapes can be found in hundreds of papers over several decades. Nobody who understands this tuning curve shape can argue against the fact of the TW.
> Indeed, the Russell & Nilsen paper is of interest in presenting
evidence against the TW theory. Figs 1D and 2B show that > the TW propagates
_beyond_ CF. That is, the authors see a response to a 15 kHz tone at positions
on the BM
> corresponding to CFs as low as 13 kHz.
But Andrew, this is what we see in any tuning curve. The term "tuning
curve" means "curve", that is there is a response besides the peak as well.
It is either the same response but at higher sound levels, or a smaller
at the same sound level. In "your" Fig. 1D the response at 90 dB goes down from 30 nm at 16 kHz (the passive BM's CF) to 1 nm at 13 kHz. This is a factor of 30. The corresponding longitudinal range on the BM is 0.9 mm. It indicates how
broadly tuned the passive BM is. We knew that for many decades. To say, on the basis of these well known data, that the TW passes the point of the BM's CF apicalwards is, sorry, not a sign of having understood the literature.
> Even more strangely, at high SPLs (100 dB) where the partition would
be expected to respond passively (a la Bekesy), we > see (Fig. 1D) that
the amplitude of response to the 15 kHz tone is more or less _constant_
from the 27-kHz region to the
> 14.5-kHz region - not what one expects from a traveling wave envelope at all.
There is nothing "strange" with the data. But a lot seems to be strange
with the way you are reading. The response to the 100 dB stimuli shows
little variation basalwards of the BM's CF, simply because it is saturated
at all measuring points. If you want to see the BM's CF in this Fig. you
have to look at the curves for 70, 80 and 90 dB. These curves show like
other tuning curves that (a) there is a TW (because of the "tail"), and (b) where about the passive BM's CF lies, which is visible at high levels.