Yahoo Tuning Groups Ultimate Backup tuning Strawberry Jam and Psychoacoustics

--- Mike Battaglia <battaglia01@...> wrote:

> "Skilled arguers, however, are not after the truth but after
> arguments supporting their views."

The piece mentions several different research programs.
I shared it because I found the bit about innate preferences
being altered by the survey paradigm interesting (though Chris
is right that 0.55 isn't much to write home of). It differs
to what Malcolm Gladwell reports on in Blink, where elaborate
surveys had no impact on people's initial choices.

I emphatically disagree with the overall interpretation on
offer; that human reasoning is no good. This is a disturbing
trend in the philosophy of science recently, e.g. the way
"behavioral economics" has been promoted in the wake of the
financapocalypse. Yesterday in the Atlantic somebody was
dissing expert knowledge. In fact, that's how I came by this
piece: http://news.ycombinator.com/item?id=2405260

-Carl

🔗Mike Battaglia <battaglia01@...>

On Tue, Apr 5, 2011 at 5:11 AM, Carl Lumma <carl@...> wrote:
>
> --- Mike Battaglia <battaglia01@...> wrote:
>
> > "Skilled arguers, however, are not after the truth but after
> > arguments supporting their views."
>
> The piece mentions several different research programs.
> I shared it because I found the bit about innate preferences
> being altered by the survey paradigm interesting (though Chris
> is right that 0.55 isn't much to write home of). It differs
> to what Malcolm Gladwell reports on in Blink, where elaborate
> surveys had no impact on people's initial choices.

I would expect that people who really want to push a certain model,
and who have the ear training necessary to identify chords, will
either consciously or subconsciously rate the chords that agree with
their model. Listening tests that don't go outside of the tuning list
are silly. Listening tests that stay exclusively in populations only
exposed to 12-tet are also silly. Listening tests that actually
purport to measure "what people like" are a little silly in and of
themselves, but may have some practical use. Listening tests that
measure "what people like" in a population that has an agenda of
pushing theories about what people like are silly.

> I emphatically disagree with the overall interpretation on
> offer; that human reasoning is no good. This is a disturbing
> trend in the philosophy of science recently, e.g. the way
> "behavioral economics" has been promoted in the wake of the
> financapocalypse. Yesterday in the Atlantic somebody was
> dissing expert knowledge. In fact, that's how I came by this
> piece: http://news.ycombinator.com/item?id=2405260

We already know the cure for confirmation bias and the placebo effect
of all kinds, and it's peer review. Although I think this is a little
harsh:

"If you're in the top 1% of the population by intelligence, curiosity,
and diligence (e.g. Richard Feynman) you will occasionally be able to
scoop experts. Everyone else... should probably not be told at a young
age that their opinions matter ex novo."

In other news, I respectfully ask that you respond to my message from
a few days ago offlist, where it seemed like we were actually starting
to make some progress. I'm a slight hypocrite for not having finished
the Continuous Harmonic Spaces pdf yet, but I'm working on it.

-Mike

🔗Carl Lumma <carl@...>

--- Mike Battaglia <battaglia01@...> wrote:

> I would expect that people who really want to push a certain
> model, and who have the ear training necessary to identify
> chords, will either consciously or subconsciously rate the
> chords that agree with their model. Listening tests that don't
> go outside of the tuning list are silly.

How many people here can identify enough xenharmonic chords
in a blinded study to sway the results? Not many. By the
way, the listening test you completed did not have 26:30:39,
7:9:11, 26:30:39, or 18:22:27 in it.

> Listening tests that stay exclusively in populations only
> exposed to 12-tet are also silly.

Looks like we have to take up interstellar travel to
do these.

> We already know the cure for confirmation bias and the placebo
> effect of all kinds, and it's peer review.

I think you mean experimental controls.

> In other news, I respectfully ask that you respond to my
> message from a few days ago offlist, where it seemed like we
> were actually starting to make some progress. I'm a slight
> hypocrite for not having finished the Continuous Harmonic
> Spaces pdf yet, but I'm working on it.

Ping me offlist with which one you mean.

-Carl

🔗Mike Battaglia <battaglia01@...>

On Tue, Apr 5, 2011 at 6:05 AM, Carl Lumma <carl@...> wrote:
>
> --- Mike Battaglia <battaglia01@...> wrote:
>
> > I would expect that people who really want to push a certain
> > model, and who have the ear training necessary to identify
> > chords, will either consciously or subconsciously rate the
> > chords that agree with their model. Listening tests that don't
> > go outside of the tuning list are silly.
>
> How many people here can identify enough xenharmonic chords
> in a blinded study to sway the results? Not many. By the
> way, the listening test you completed did not have 26:30:39,
> 7:9:11, 26:30:39, or 18:22:27 in it.

But I thought that they did, which may have biased the results
slightly in favor of what I thought.

> > Listening tests that stay exclusively in populations only
> > exposed to 12-tet are also silly.
>
> Looks like we have to take up interstellar travel to
> do these.

> > We already know the cure for confirmation bias and the placebo
> > effect of all kinds, and it's peer review.
>
> I think you mean experimental controls.

No, I mean peer review. The greater philosophical point is that
different biases are failures of reason. How can we reason our way to
enlightenment when we our faculties of reason might be influenced by a
bias that we don't know of? Peer review solves these problems.

There are plenty of experiments that have controls but still fall prey
to some sort of methodological bias. Other researchers who don't
share your bias pick up on this and point it out. And life goes on.

-Mike

🔗Carl Lumma <carl@...>

-- Mike Battaglia <battaglia01@...> wrote:

> > How many people here can identify enough xenharmonic chords
> > in a blinded study to sway the results? Not many. By the
> > way, the listening test you completed did not have 26:30:39,
> > 7:9:11, 26:30:39, or 18:22:27 in it.
>
> But I thought that they did, which may have biased the results
> slightly in favor of what I thought.

Right, but unless people are getting it right, these biases
will average out.

> > > Listening tests that stay exclusively in populations only
> > > exposed to 12-tet are also silly.
> >
> > Looks like we have to take up interstellar travel to

My understanding is that everybody's now getting a majority
of 12-ET (or nearly 12-ET) from a young age (except perhaps a
few isolated tribes in Africa, Papua New Guinea, the Amazon).
I mean, I'm sure there are plenty of Indian, Arab etc musicians
with kids who are significantly exposed. But I think you'd be
hard pressed to find an indigenous form which is still dominant
among the general population. I could be wrong. Perhaps any
musicologists out there will chime in.

> > > We already know the cure for confirmation bias and the
> > > placebo effect of all kinds, and it's peer review.
> >
> > I think you mean experimental controls.
>
> No, I mean peer review.

The primary purpose of peer review is to establish notability.
It is not for catching bias.

-Carl

🔗Mike Battaglia <battaglia01@...>

On Tue, Apr 5, 2011 at 6:33 AM, Carl Lumma <carl@...> wrote:
>
> -- Mike Battaglia <battaglia01@...> wrote:
>
> > > How many people here can identify enough xenharmonic chords
> > > in a blinded study to sway the results? Not many. By the
> > > way, the listening test you completed did not have 26:30:39,
> > > 7:9:11, 26:30:39, or 18:22:27 in it.
> >
> > But I thought that they did, which may have biased the results
> > slightly in favor of what I thought.
>
> Right, but unless people are getting it right, these biases
> will average out.

OK, but if the tests are limited to the tuning list, or if a chord
sounds similar enough to another chord that a large group of people
hear it as such, this may not happen.

> > > Looks like we have to take up interstellar travel to
>
> My understanding is that everybody's now getting a majority
> of 12-ET (or nearly 12-ET) from a young age (except perhaps a
> few isolated tribes in Africa, Papua New Guinea, the Amazon).
> I mean, I'm sure there are plenty of Indian, Arab etc musicians
> with kids who are significantly exposed. But I think you'd be
> hard pressed to find an indigenous form which is still dominant
> among the general population. I could be wrong. Perhaps any
> musicologists out there will chime in.

I didn't say that it was necessary to only explore populations that
had no exposure to 12-TET. A population that has been exposed to
12-TET but also a second tuning system might be a decent choice.

One can also look for psychoacoustic invariables by exploring what
they are not: research among expert musicians from different tuning
systems, for example, would be really valuable by showing how their
perception differs. Anything that differs is, then, not a
psychoacoustically hard-wired facet of music cognition. Things that
are common may still not be psychoacoustically hard-wired either, so
you can get false positives from this, but never false negatives.

> > > > We already know the cure for confirmation bias and the
> > > > placebo effect of all kinds, and it's peer review.
> > >
> > > I think you mean experimental controls.
> >
> > No, I mean peer review.
>
> The primary purpose of peer review is to establish notability.
> It is not for catching bias.

I guess peer review's not really it. Publication was more what I was after.

-Mike

🔗Michael <djtrancendance@...>

>"Listening tests that don't go outside of the tuning list are silly. Listening tests that stay exclusively in populations only exposed to 12-tet are also silly."

Agreed. The thing the article seems to indirectly say, though, is that we should follow our EARS rather than get wrapped up in numerical logic and that by following EARS results of the common folk and experts (provided experts are not provided with incentive for bias) could coincide. If numerical logic MIRRORS what our ears think, though...we may be onto something. Far as what kind of populations to test on...I think it's fair to take people exposed to 12TET and introduce them over a few weeks to scales/music from other cultures and, only after that, test microtonal theories on them.

🔗Carl Lumma <carl@...>

--- Mike Battaglia <battaglia01@...> wrote:

> > > > > We already know the cure for confirmation bias and the
> > > > > placebo effect of all kinds, and it's peer review.
> > > >
> > > > I think you mean experimental controls.
> > >
> > > No, I mean peer review.
> >
> > The primary purpose of peer review is to establish notability.
> > It is not for catching bias.
>
> I guess peer review's not really it. Publication was more what
> I was after.

Controls are what you're after, including repetition of
results. Publication helps with the latter. -C.

🔗lobawad <lobawad@...>

--- In tuning@yahoogroups.com, "Carl Lumma" <carl@...> wrote:
>
> --- Mike Battaglia <battaglia01@> wrote:
>
> > "Skilled arguers, however, are not after the truth but after
> > arguments supporting their views."
>
> The piece mentions several different research programs.
> I shared it because I found the bit about innate preferences
> being altered by the survey paradigm interesting (though Chris
> is right that 0.55 isn't much to write home of).

The preferences which were altered were not "innate". These were college students being tested, not infants.

🔗lobawad <lobawad@...>

--- In tuning@yahoogroups.com, Mike Battaglia <battaglia01@...> wrote:
I think you mean experimental controls.
>
> No, I mean peer review. The greater philosophical point is that
> different biases are failures of reason. How can we reason our way to
> enlightenment when we our faculties of reason might be influenced by a
> bias that we don't know of? Peer review solves these problems.
>
> There are plenty of experiments that have controls but still fall prey
> to some sort of methodological bias. Other researchers who don't
> share your bias pick up on this and point it out. And life goes on.
>
> -Mike
>

A cynic would say that one of things that makes "peers" your peers is that they share your biases.

Strawberry Jam and Psychoacoustics

🔗Carl Lumma <carl@...>

🔗Chris Vaisvil <chrisvaisvil@...>

🔗Mike Battaglia <battaglia01@...>

🔗Carl Lumma <carl@...>

🔗Mike Battaglia <battaglia01@...>

🔗Carl Lumma <carl@...>

🔗Mike Battaglia <battaglia01@...>

🔗Carl Lumma <carl@...>

🔗Mike Battaglia <battaglia01@...>

🔗Michael <djtrancendance@...>

🔗Carl Lumma <carl@...>

🔗lobawad <lobawad@...>

🔗lobawad <lobawad@...>