Subject: Re: [AUDITORY] Long/Low & Short/High? From: Sarah Hawkins <sh110@xxxxxxxx> Date: Sun, 6 Nov 2022 15:58:25 +0000--------------XrsHhBid0f4B7XOapV7hsbk2 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by edgeum3.it.mcgill.ca id 2A6FwYB6006836 Dear Lori I add an example from speech perception to those other people have sent=20 for music. Noel Nguyen and I found an association between pitch and perceived stop=20 voicing in synthetic stimuli. Importantly from your point of view, the=20 difference was found in perception but not in production. I explain at=20 some length because it's a bit complicated and not that well known: the=20 stop in question (/d/ or /t/) was at the end of a word whereas the=20 manipulated pitch difference was at the beginning. =C2=A0 We (and, separately, John Coleman) examined cues to phonological=20 voicing contrast in word-final /d/ vs /t/ as in led vs let. Those cues=20 are distributed over the entire syllable, including having longer /l/ as=20 well as the better-known longer vowel. There were also systematic=20 spectral differences in the [l]. No differences in f0 in this natural=20 lab speech though. =C2=A0=C2=A0=C2=A0=C2=A0 Hawkins, S., & Nguyen, N. (2004). Influence of = syllable-coda=20 voicing on the acoustic properties of syllable-onset /l/ in English.=20 Journal of Phonetics, 32(2), 199-231. In a series of perceptual experiments, we manipulated durational and=20 spectral cues to coda voicing in the onset of high-quality synthetic led=20 vs let. We manipulated f0 as an independent variable because we had to=20 make the usual types of decision re what to do with it when the duration=20 of the [l] varied systematically and the pitch contour was to be a=20 natural-sounding declination. Listeners heard only the [l] and the first=20 few (80?) ms of the vowel. They had to respond (i.e. guess) whether the=20 word was led or let. Duration of [l] was a strong cue but so was f0. Strong enough for us to=20 have noticed it when making the stimuli, which is another reason we made=20 it an IV. There were some interesting learning effects as well, but the point here=20 is that English (and probably French) listeners may use low pitch as a=20 cue to phonological voicing throughout the syllable. As you know, this=20 difference in f0 is consistent with what happens around closure for the=20 word-final stop itself, but as far as I know there's not evidence for it=20 in the word onset in natural speech -- just for the durational and=20 spectral difference. =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 Hawkins, S., & Nguyen, N. (2001). P= erception of coda voicing=20 from properties of the onset and nucleus of 'led' and 'let'. In P.=20 Dalsgaard, B. Lindberg, & H. Benner (Eds.), Eurospeech 2001: Proceedings=20 of the 7th International Conference on Speech Communication and=20 Technology, (Vol. 1, pp. 407-410). As a general comment, it may be worth considering the effect as related=20 to onomatopoeic tendencies. There are always exceptions, but as far as I=20 know there is some degree of association between size and pitch: little=20 things higher and 'smaller' (shorter, faster, etc), big things lower and=20 clunkier.=C2=A0 Noel and I suggested that, rather than the standard term = coda=20 'voicing', it might be better to describe the whole syllable as 'sombre'=20 for words like /led./ and 'bright' for words like /let./ This seems in=20 line with both the whole-word complexity of the word-final voicing=20 distinction in English, and the possibility that pitch can influence=20 perception metaphorically, as it were. all best Sarah On 01/11/2022 20:00, Lori Holt wrote: > > Dear auditory aficionados, > > Fred Dick and I have been doing some work with a long-short tone=20 > duration identification task (50ms vs 90ms) where tone frequency is=20 > chosen from one of x values in a truncated range (e.g., 800Hz, 920Hz,=20 > 1000Hz, 1080Hz, 1200Hz).=C2=A0=C2=A0(You might be familiar with this pa= radigm=20 > from Mondor & Bregman, 1994). > > We have found a weak but quite reliable effect, where subjects tend to=20 > judge lower frequencies more often as 'long', and higher frequencies=20 > as 'short'.=C2=A0 This was unexpected yet remarkably consistent across = a=20 > lot of experiments. We have been unable to track downmention of this=20 > in the literature. > > We did dig up a few papers that purported to be on the general topic=20 > of frequency effects on duration judgments, but these ended up being=20 > about different things entirely... > > We wondered whether anyone might be familiar with literature we've=20 > missed- or maybe even have encountered something like this before=20 > yourself? > > Best wishes, > Lori & Fred > > > ______________________________________________________________ > loriholt > Professor | Department of Psychology > Professor | Neuroscience Institute > Co-Director | Center for the Neural Basis of Cognition > Co-Director | Behavioral Brain (B2) Research Training Program (NIGMS) > Carnegie Mellon University > loriholt@xxxxxxxx=C2=A0| www.psy.cmu.edu/~holtlab/=20 > <http://www.psy.cmu.edu/~holtlab/> > pronouns: she, her, hers > --------------XrsHhBid0f4B7XOapV7hsbk2 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: 8bit <html><head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> </head> <body> Dear Lori<br> <br> I add an example from speech perception to those other people have sent for music.<br> <br> Noel Nguyen and I found an association between pitch and perceived stop voicing in synthetic stimuli. Importantly from your point of view, the difference was found in perception but not in production. I explain at some length because it's a bit complicated and not that well known: the stop in question (/d/ or /t/) was at the end of a word whereas the manipulated pitch difference was at the beginning.<br> We (and, separately, John Coleman) examined cues to phonological voicing contrast in word-final /d/ vs /t/ as in led vs let. Those cues are distributed over the entire syllable, including having longer /l/ as well as the better-known longer vowel. There were also systematic spectral differences in the [l]. No differences in f0 in this natural lab speech though.<br> Hawkins, S., & Nguyen, N. (2004). Influence of syllable-coda voicing on the acoustic properties of syllable-onset /l/ in English. Journal of Phonetics, 32(2), 199-231. <br> <br> In a series of perceptual experiments, we manipulated durational and spectral cues to coda voicing in the onset of high-quality synthetic led vs let. We manipulated f0 as an independent variable because we had to make the usual types of decision re what to do with it when the duration of the [l] varied systematically and the pitch contour was to be a natural-sounding declination. Listeners heard only the [l] and the first few (80?) ms of the vowel. They had to respond (i.e. guess) whether the word was led or let. <br> Duration of [l] was a strong cue but so was f0. Strong enough for us to have noticed it when making the stimuli, which is another reason we made it an IV. <br> <br> There were some interesting learning effects as well, but the point here is that English (and probably French) listeners may use low pitch as a cue to phonological voicing throughout the syllable. As you know, this difference in f0 is consistent with what happens around closure for the word-final stop itself, but as far as I know there's not evidence for it in the word onset in natural speech -- just for the durational and spectral difference.<br> Hawkins, S., & Nguyen, N. (2001). Perception of coda voicing from properties of the onset and nucleus of 'led' and 'let'. In P. Dalsgaard, B. Lindberg, & H. Benner (Eds.), Eurospeech 2001: Proceedings of the 7th International Conference on Speech Communication and Technology, (Vol. 1, pp. 407-410). <br> <br> As a general comment, it may be worth considering the effect as related to onomatopoeic tendencies. There are always exceptions, but as far as I know there is some degree of association between size and pitch: little things higher and 'smaller' (shorter, faster, etc), big things lower and clunkier. Noel and I suggested that, rather than the standard term coda 'voicing', it might be better to describe the whole syllable as 'sombre' for words like <i>led.</i> and 'bright' for words like <i>let.</i> This seems in line with both the whole-word complexity of the word-final voicing distinction in English, and the possibility that pitch can influence perception metaphorically, as it were.<br> <br> all best<br> Sarah<br> <br> <br> <div class="moz-cite-prefix">On 01/11/2022 20:00, Lori Holt wrote:<br> </div> <blockquote type="cite" cite="mid:CADMGEwMhF4zf9FbyqfvBJGKQzuOh3aGPbDWLe4w4WZ+K0EqxnQ@xxxxxxxx"> <div dir="ltr"> <div class="gmail_default" style="font-family:verdana,sans-serif"><br clear="all"> </div> <div> <div dir="ltr" class="gmail_signature" data-smartmail="gmail_signature"> <div dir="ltr"> <div style="font-size:12.8px"> <div style="font-family:verdana,sans-serif" class="gmail_default">Dear auditory aficionados,</div> <div style="font-family:verdana,sans-serif" class="gmail_default"><br> </div> <div style="font-family:verdana,sans-serif" class="gmail_default"> <div><font size="2"><span style="font-family:verdana,sans-serif">F<span class="gmail_default">red Dick</span> and I have been doing some work w<span class="gmail_default" style="font-family:verdana,sans-serif"></span>ith a long-short tone duration identification task (50ms vs 90ms) where tone frequency is chosen from one of x values in a truncated range (e.g., 800Hz, 920Hz, 1000Hz, 1080Hz, 1200Hz). (You might be familiar with this paradigm from Mondor & Bregman, 1994). </span></font></div> <div><font size="2"><span style="font-family:verdana,sans-serif"></span></font></div> <div> <div> <div dir="ltr"> <div dir="ltr"> <div style="font-size:12.8px"><font size="2"><span style="font-family:verdana,sans-serif"> <br> We have found a weak but quite reliable effect, where subjects tend to judge lower frequencies more often as 'long', and higher frequencies as 'short'. This was unexpected yet remarkably consistent across a lot of experiments. We have been unable to track down<span class="gmail_default" style="font-family:verdana,sans-serif"> </span>mention of this in the literature. <br> <br> We did dig up a few papers that purported to be on the general topic of frequency effects on duration judgments, but these ended up being about different things entirely... <br> <br> <span class="gmail_default" style="font-family:verdana,sans-serif">We wondered whether</span> anyone might be familiar with literature we've missed- or maybe even have encountered something like this before yourself? <br> </span></font></div> <div style="font-size:12.8px"><font size="2"><span style="font-family:verdana,sans-serif"><br> </span></font></div> <div style="font-size:12.8px"><font size="2"><span style="font-family:verdana,sans-serif"><span class="gmail_default" style="font-family:verdana,sans-serif">Best wishes,</span></span></font></div> <div style="font-size:12.8px"><font size="2"><span style="font-family:verdana,sans-serif"><span class="gmail_default" style="font-family:verdana,sans-serif">Lori & Fred</span></span></font></div> </div> </div> </div> </div> </div> <br> </div> <div style="font-size:12.8px"><br> </div> <div style="font-size:12.8px">______________________________________________________________</div> <span style="font-family:"arial black",sans-serif;font-size:large"><font color="#999999">lori</font></span><span style="font-family:"arial black",sans-serif;font-size:large"><font color="#444444">holt</font></span> <div> <div style="font-size:small"><font color="#999999"><font face="tahoma, sans-serif">Professor </font></font><font color="#999999"><font face="tahoma, sans-serif"><font color="#999999"><span style="font-family:tahoma,sans-serif">| </span></font>Department of Psychology</font></font></div> <div style="font-size:small"><font color="#999999"><span style="font-family:tahoma,sans-serif">Professor | Neuroscience Institute <br> </span></font></div> <div style="font-size:small"><font color="#999999"><span style="font-family:tahoma,sans-serif">Co-Director | Center for the Neural Basis of Cognition<br> </span></font></div> <div style="font-size:small"><font color="#999999"><span style="font-family:tahoma,sans-serif">Co-Director | Behavioral Brain (B2) Research Training Program (NIGMS)<br> </span></font></div> <div style="font-size:small"><font face="tahoma, sans-serif" color="#999999">Carnegie Mellon University</font></div> <div style="font-size:small"><span style="color:rgb(102,102,102)"><font face="tahoma, sans-serif"><a href="mailto:loriholt@xxxxxxxx" target="_blank" moz-do-not-send="true" class="moz-txt-link-freetext">loriholt@xxxxxxxx</a> | </font><font face="tahoma, sans-serif"><a href="http://www.psy.cmu.edu/~holtlab/" target="_blank" moz-do-not-send="true">www.psy.cmu.edu/~holtlab/</a></font></span></div> </div> <font color="#999999"><font size="1">pronouns: she, her, hers</font></font> <div><br> </div> </div> </div> </div> </div> </blockquote> <br> </body> </html> --------------XrsHhBid0f4B7XOapV7hsbk2--