[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index] [Thread Index] [Author Index]
Return to main CEDA-L Archive Page

answering Larson

Howdy Professor Larson!!!  I respect Professor Gary Larson tremendously.  He 
ain't right about the research I post, though.  Well, the nasty bits are 
toward the end if you're just browsing for cheap thrills.

He says that 250 wpm "might not square with competitive norms."

Well, 250 wpm is the optimal comprehension rate for folks not previously 
exposed to speedy speech.  The moral to draw is that debaters ought to slow 
down for lay audiences and critics (Yes, you ought to say that you heard 
that HERE first).  Debate audiences and critics usually prefer and 
comprehend optimally at much higher rates (they're "trained") so folks ought 
to adapt appropriately.  Also, comprehension isn't the only valid 
pedagogical, strategic, or entertainment concern (some mis-comprehension is 
usually a nice touch).  As far as being able to merely REMEMBER the speech, 
for example, the final round was cozy and pokey.  Professors Wingfield, 
Poon, Lombardi, and Lowe report in the 1985 Journal of Gerontology that:  
"Both our young and elderly participants showed excellent recall for five- 
and eight-word normal sentences even when speech rates reached a level of 
425 wpm, more than twice the rate at which normal conversation is ordinarily 
heard." (p583)

He says that training gets folks to 275 wpm, still too slow for The Buzz.  
(although plenty for The Biz if Lemoine's retirement conversion is to be 
believed - Dave, don't let him slow her down!)

Well, the Army training and exposure studies were for relatively brief 
periods of time (less than 24 total exposure hours at the most), used army 
personnel (no offense to the folks at the USMA), and essentially consisted 
of giving extra leave time for anyone who did better faster.  Debaters and 
critics get much longer exposures from a greater variety of sources over a 
longer period of time, are usually more highly motivated to learn, ain't 
enlisted as targets, and have topic familiarity, flowing, and debater 
personal attention to aid in "training."  No research  to cite that says 
debate folks do better, but the research certainly does not establish 275 
wpm as the best that we can be.  The Foulke remarks about blind college 
students ONLY reports PREFERRED rates of speech, not the rates at which they 
comprehend optimally - we can only guess what that is - I guess that it is 
786 wpm.

He says that compressed speech is "cleaner" than speedy speech.

Well, I think that he is making this up.  

1)  To my knowledge there haven't been any studies which say this (I have 
looked) and Professor Larson follows the tradition of not citing any sources 
at all.  Almost no one has compared naturally speedy speech against 
compressed speech with respect to comprehensibility.  Certainly no one has 
compared the speedy speech of debaters (TRAINED to be speedy AND 
comprehensible) to compressed speech with respect to comprehensibility.  
That means he is just making it up.  What research we have is cautious but 
warns against the LOWER QUALITY OF COMPRESSED SPEECH.  Professors Stine, 
Wingfield, and Poon, for example, note in their 1986 article entitled "How 
Much and How Fast:  Rapid Processing of Spoken Language in Later Adulthood" 
in the journal Psychology and Aging:  "The
extent to which the processing rate differences were exacerbated by a loss 
of signal richness associated with time compression is important, especially 
in the context of auditory discrimination deficits that often accompany 
normal aging (Corso, 1984).  In general, older adults react to naturally 
spoken rapid speech in a similar way to artificially time-compressed speech 
of equivalent speech rates (Schmitt & Carroll, 1985), but the possibility of 
multiple sources of a processing deficit must be kept in mind." (p306)

2)  The "explanation" of why compressed speech is "cleaner" is based on a 
misunderstanding of compression.  Compression doesn't "reduce space between 
phonetic segments."  That sounds like it just gets rid of empty and useless 
silence.  What happens is that a reading is taped at normal speed and the 
tape is played back at a multiplied speed:  but, because that would be 
incomprehensible (you know what a tape at double speed sounds like - a high 
pitched squeal), RANDOMLY selected portions of the tape are omitted at 
playback which lowers the pitch of playback to normal frequencies.  Now, how 
randomly skipping parts of a recorded text (phonemes, segments, whatever 
happens to be on that bit of tape) is supposed to be "cleaner" than merely 
saying the whole text quickly and at a normal pitch, I have no idea - that 
is what is referred to above as a loss of "signal richness."  Also, remember 
that the faster the compressed speech, the more has to be randomly selected 
out and the more information is lost.  Professors Stine, Wingfield, and Poon 
briefly explain how a Lexicon Varispeech II compressor/expander (state o' 
the art tech) does it:   "... small segments (20 ms in length) are 
periodically deleted, and the remaining segments are abutted in time.  The 
result is speech reproduced in less than normal time, but without 
substantial distortion in pitch or quality that would, for example, 
accompany tape-recorder playback at faster than normal speed.  The degree of 
time compression is controlled by the frequency with which tape segments are 
deleted." (p305).  Now, the AmBiChron speech processor (the old tech)  
sounds even worse than the Lexicon Varispeech II.  Gade et al explain in 
Human Learning of 1984:  "Although no entire speech sound is likely to be 
lost completely because the interval is generally shorter than any single 
speech sound, some critical features may at times be deleted.  Although 
pitch-normalized output tends to preserve phrasings, stress, and pauses in 
the original speech record, increases above twice the normal speed tend to 
produce some distortions." (p97)

3)  No research has been done on this, but these are my ad-hoc, 
make-it-up-as-you-go, "if there's no ev say it anyway - they might believe 
you", Top Ten Reasons Debaters Are More Comprehensible Than The Lexicon 
Varispeech II at high rates of speech:  1) eye contact, facial expression, 
gestures, and in-person presence 2) natural adjustment of pitch without loss 
of signal richness 3) they do the 
Northwestern/Pittsburgh/Madsen/Dauber/Korcok/Devereaux/John Dean the King 
"speaking" exercises better than the tape recorder does them 4) by looking 
at their critic they get a clue about incomprehension and repeat, redund, 
recast, or explain 5) they train, practice, and perfect the specific arts of 
speedy speech 6) debate critics try harder to understand debaters than they 
do Varispeech IIs 7) they intelligently hilite cards to get rid of confusing 
verbiage and obfuscating words 8)  they do comprehension tricks unknown to 
tape recorders like start slow for the first 15 seconds, slow down for the 
payoff sentences, lower their pitch more for the good bits, etc.  9)  they 
tell stories, use tags, refer to the flow, use the extensive contextual 
knowledge of the critics, etc. to remain comprehensible while doing THE BUZZ 
and 10) just because.

Professor Larson then says a number of true and important things about the 
importance of well-articulated speech at any speed.  I agree.  Remember, I 
debated my senior year with Kevin O'Leary.  His cross-ex answers were 
invariably "mmmphnblmbd" -- on the rare occasion that someone asked him to 
stop mumbling and to clarify he wouold look up, look right at the critic, 
get the hair out of his eyes, and say "mmphphnblnblmbd."  Professor Stepp 
loved it and several times rewarded it with 20 speaker points - she liked ME 
(I got 28s and 29s).  Although John Dean and Bill Boggs used to get 30s from 
her - pissed me off...

:( michael korcok :)
Archive created by Jonathan Stanton (jonathan@cs.jhu.edu)
Return to main CEDA-L Archive Page