# Attaching labels

**URL:** <http://discourse.iapct.org/t/attaching-labels/1349>\
**Category:** CSG2004\
**Created:** [October 30, 2004, 5:31pm UTC](http://discourse.iapct.org/t/attaching-labels/1349 "2004-10-30T17:31:49Z")\
**Posts on this page:** 20\
**Page:** 1

<div class="post-metadata">

**Author:** ![Richard\_Marken2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/r/edb3f5/32.png) [@Richard\_Marken2](http://discourse.iapct.org/u/Richard_Marken2)\
**Post date:** [October 30, 2004, 5:31pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/1 "2004-10-30T17:31:49Z")

</div>

[From Rick Marken (2004.10.30.1030)]

> Martin Taylor (2004.10.30.11.50) --

> The question then is whether, within the general PCT framework, the  
> following choices are the only possible ones or whether there are  
> other possibilities: the labelling experience shows (1) that the  
> structure of perceptual functions is not a strict hierarchy, or (2)  
> that labelling is itself a control process in which the action of  
> some output is not through lower-level outputs on the outer world but  
> is directly on the functioning of the perceptual part of the control  
> hierarchy.
> 
> ==============end question============
> 
> Now I ask Bruce whether this is a fair statement of his question, and  
> I ask everyone whether it is a reasonably stated question in PCT.

I think that the "labeling experience", per se, could suggest  
hypotheses about the structure of perceptual functions and about the  
nature of certain control processes. But I don't think that simply  
noticing that people can call a spade a "spade" (and a club a "club"  
and so on) -- that is, that people can label aspects of their  
experience -- tells you much about what they are doing (whether they  
are responding or controlling, for example) or how they are doing it  
(whether or not it's a hierarchical control process, for example). I  
think you need to set up a carefully designed situation which both  
defines what you mean by "labeling" and gives you the ability to  
manipulate variables and develop models that let you figure out what  
might be going on when people attach labels.

Best

Rick

> **···**
>
> ---  
> Richard S. Marken  
> marken@mindreadings.com  
> Home 310 474-0313  
> Cell 310 729-1400

---

<div class="post-metadata">

**Author:** ![Martin\_Taylor2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/m/b4bc9f/32.png) [@Martin\_Taylor2](http://discourse.iapct.org/u/Martin_Taylor2)\
**Post date:** [October 30, 2004, 6:02pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/2 "2004-10-30T18:02:54Z")

</div>

[Martin Taylor 2004.10.30.13.59]

> [From Rick Marken (2004.10.30.1030)]
> 
> > Martin Taylor (2004.10.30.11.50) --
> 
> > The question then is whether, within the general PCT framework, the  
> > following choices are the only possible ones or whether there are  
> > other possibilities: the labelling experience shows (1) that the  
> > structure of perceptual functions is not a strict hierarchy, or (2)  
> > that labelling is itself a control process in which the action of  
> > some output is not through lower-level outputs on the outer world but  
> > is directly on the functioning of the perceptual part of the control  
> > hierarchy.
> > 
> > ==============end question============
> > 
> > Now I ask Bruce whether this is a fair statement of his question, and  
> > I ask everyone whether it is a reasonably stated question in PCT.
> 
> I think that the "labeling experience", per se, could suggest  
> hypotheses about the structure of perceptual functions and about the  
> nature of certain control processes. But I don't think that simply  
> noticing that people can call a spade a "spade" (and a club a "club"  
> and so on) -- that is, that people can label aspects of their  
> experience -- tells you much about what they are doing (whether they  
> are responding or controlling, for example) or how they are doing it  
> (whether or not it's a hierarchical control process, for example). I  
> think you need to set up a carefully designed situation which both  
> defines what you mean by "labeling" and gives you the ability to  
> manipulate variables and develop models that let you figure out what  
> might be going on when people attach labels.

Quite so, as far as you go. But you missed the critical element of  
the question as related to experience, which is that not only can the  
perception of label be evoked by seeing the object, but also the  
visual perception of the object can be evoked by hearing (or reading)  
the label. I think a lot of people would agree that both experiences  
happen.

It's the bidirectionality of the experience that raises the question  
about the possible structural relations among perceptual functions.

Martin

---

<div class="post-metadata">

**Author:** ![Richard\_Marken2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/r/edb3f5/32.png) [@Richard\_Marken2](http://discourse.iapct.org/u/Richard_Marken2)\
**Post date:** [October 31, 2004, 6:42am UTC](http://discourse.iapct.org/t/attaching-labels/1349/3 "2004-10-31T06:42:09Z")

</div>

[From Rick Marken (2004.10.30.2340)]

> Martin Taylor (2004.10.30.13.59) --

> not only can the perception of label be evoked by seeing the object,  
> but also the visual perception of the object can be evoked by hearing  
> (or reading) the label. I think a lot of people would agree that both  
> experiences happen.
> 
> It's the bidirectionality of the experience that raises the question  
> about the possible structural relations among perceptual functions.

I think you may have a point. I think I can suggest an experiment that  
would test a hypothesis about some structural relations assumed by PCT.  
&nbsp;&nbsp;The experiment would test to see if the bidirectionality you describe  
is temporally symmetrical. I think PCT would predict temporal  
asymmetry. That is, I think it should take longer to evoke the label  
given the object (instance) than to invoke the object given the label.  
To be specific, it should take longer to select "river" (the category  
label) when shown a picture of a river (rather than a mountain, say)  
than to select the picture of a river when shown the category label  
"river". My prediction is based on the assumption that the category  
level of control is higher and, thus, slower, than the configuration  
level, according to the PCT model. Selecting "river" when shown a  
picture of a river requires control at the category level. That is, you  
have to select "river" as the appropriate reference category for the  
river configuration in the picture. On the other hand, selecting the  
picture of a river given the reference category "river" requires  
control at the configuration level -- a lower and, thus, a faster level  
of control.

I have a feeling that people may already have done experiments of this  
kind so we may have data pertinent to this question of structure  
already in the literature. I can't think of the relevant studies right  
off hand. Perhaps you can.

Regards

Rick

> **···**
>
> ---  
> Richard S. Marken  
> marken@mindreadings.com  
> Home 310 474-0313  
> Cell 310 729-1400

---

<div class="post-metadata">

**Author:** ![Martin\_Taylor2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/m/b4bc9f/32.png) [@Martin\_Taylor2](http://discourse.iapct.org/u/Martin_Taylor2)\
**Post date:** [October 31, 2004, 2:58pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/4 "2004-10-31T14:58:00Z")

</div>

> [From Rick Marken (2004.10.30.2340)]
> 
> > Martin Taylor (2004.10.30.13.59) --
> 
> > not only can the perception of label be evoked by seeing the object,  
> > but also the visual perception of the object can be evoked by hearing  
> > (or reading) the label. I think a lot of people would agree that both  
> > experiences happen.
> > 
> > It's the bidirectionality of the experience that raises the question  
> > about the possible structural relations among perceptual functions.
> 
> I think you may have a point. I think I can suggest an experiment that  
> would test a hypothesis about some structural relations assumed by PCT.  
> The experiment would test to see if the bidirectionality you describe  
> is temporally symmetrical. I think PCT would predict temporal  
> asymmetry. That is, I think it should take longer to evoke the label  
> given the object (instance) than to invoke the object given the label.  
> To be specific, it should take longer to select "river" (the category  
> label) when shown a picture of a river (rather than a mountain, say)  
> than to select the picture of a river when shown the category label  
> "river". My prediction is based on the assumption that the category  
> level of control is higher and, thus, slower, than the configuration  
> level, according to the PCT model. Selecting "river" when shown a  
> picture of a river requires control at the category level. That is, you  
> have to select "river" as the appropriate reference category for the  
> river configuration in the picture. On the other hand, selecting the  
> picture of a river given the reference category "river" requires  
> control at the configuration level -- a lower and, thus, a faster level  
> of control.

This kind of experiment could be useful, but I think its  
interpretation might be ambiguous. The problem is that in the  
original case, the "secondary" perception is imaginary and the issue  
is how this imaginary perception is invoked. In your experiment,  
either both perceptions are data-based and what is being timed is how  
long it takes the subject to make and to express a choice, or the  
secondary perception is imaginary and the test is how long it takes  
the subject to express it in a way interpretable by the experimenter.

If it's a choice question (choose this picture or choose this written  
label), one has the problem of the time it takes to perceive the  
choice selection as well as how long it takes to choose, and that's  
not an easy pair to tease apart.

If it's an imagination question, clearly it's quicker to say "river"  
than to draw an acceptable picture of one.

> I have a feeling that people may already have done experiments of this  
> kind so we may have data pertinent to this question of structure  
> already in the literature. I can't think of the relevant studies right  
> off hand. Perhaps you can.

Possibly studies of picture-word Stroop might be relevant. But again  
there is the problem that both perceptions are evoked by external  
data, whereas the issue is in the nature of the feedback path(s) that  
evoke the imaginary component.

At the moment, I can't think of other relevant studies. Nor can I  
invent one. But there must be a way.

Martin

---

<div class="post-metadata">

**Author:** ![Bill\_Powers1](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/b/8e7dd6/32.png) [@Bill\_Powers1](http://discourse.iapct.org/u/Bill_Powers1)\
**Post date:** [October 31, 2004, 3:32pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/5 "2004-10-31T15:32:52Z")

</div>

[From Bill Powers (2004.10.31.0830 MST)]

Rick Marken (2004.10.30.2340) --

Good experiment. If nobody else volunteers, do it!

Best,

Bill P.

---

<div class="post-metadata">

**Author:** ![Martin\_Taylor2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/m/b4bc9f/32.png) [@Martin\_Taylor2](http://discourse.iapct.org/u/Martin_Taylor2)\
**Post date:** [October 31, 2004, 4:08pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/6 "2004-10-31T16:08:32Z")

</div>

[Martin Taylor 2004.10.31.12.04]

> [From Rick Marken (2004.10.31.0820)]
> 
> > Martin Taylor (October 31, 2004, at 06:58 AM)
> 
> > > Rick Marken (2004.10.30.2340)
> 
> > > I think you may have a point. I think I can suggest an experiment  
> > > that  
> > > would test a hypothesis about some structural relations assumed by  
> > > PCT.  
> > > The experiment would test to see if the bidirectionality you describe  
> > > is temporally symmetrical.
> > 
> > This kind of experiment could be useful, but I think its  
> > interpretation might be ambiguous...
> 
> > At the moment, I can't think of other relevant studies. Nor can I  
> > invent one. But there must be a way.
> 
> Well, you just go ahead and settle back in your armchair. I think I'll  
> take Bill's advice and do the experiment myself if I can't find  
> something similar in the literature.

That isn't really what I'm doing, literally or metaphorically. Quite  
the contrary, actually. I'm trying to keep, if not my head, at least  
my nostrils above water in the swamp of work. But CSGnet (and I wish  
ECACS) is occasionally interesting enough for me to take time off to  
try to contribute something of whatever little value I can.

> **···**
>
> --------------------------
> 
> What I normally ask for, before embarking on any experiment, is an  
> understanding of what question the experiment would be likely to  
> answer if the data were to come out clean. In the case of your  
> proposal, as I said before, I think your experiment would provide  
> interesting results, but not results that bear on the question being  
> asked, which is about the two-way evocation of imagined perceptions.  
> Since in the experiment (if I understood your proposal correctly)  
> both perceptions are being invoked by externally provided data, the  
> two-way feedback connection(s) of interest is/are, if not eliminated  
> from consideration, at least compromised.
> 
> In the context of your experiment, I suggested that results on  
> picture-word Stroop might provide useful information. Lochlan Magee,  
> perhaps 20-25 years ago, did his thesis on that. I don't know where  
> he published, but if you want me to follow up, I can ask him this  
> week. I don't imagine he's the only one who worked in that area,  
> though.
> 
> Martin

---

<div class="post-metadata">

**Author:** ![Richard\_Marken2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/r/edb3f5/32.png) [@Richard\_Marken2](http://discourse.iapct.org/u/Richard_Marken2)\
**Post date:** [October 31, 2004, 4:21pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/7 "2004-10-31T16:21:05Z")

</div>

[From Rick Marken (2004.10.31.0820)]

> Martin Taylor (October 31, 2004, at 06:58 AM)

> > Rick Marken (2004.10.30.2340)

> > I think you may have a point. I think I can suggest an experiment  
> > that  
> > would test a hypothesis about some structural relations assumed by  
> > PCT.  
> > The experiment would test to see if the bidirectionality you describe  
> > is temporally symmetrical.
> 
> This kind of experiment could be useful, but I think its  
> interpretation might be ambiguous...

> At the moment, I can't think of other relevant studies. Nor can I  
> invent one. But there must be a way.

Well, you just go ahead and settle back in your armchair. I think I'll  
take Bill's advice and do the experiment myself if I can't find  
something similar in the literature.

Best

Rick

> **···**
>
> ---  
> Richard S. Marken  
> marken@mindreadings.com  
> Home 310 474-0313  
> Cell 310 729-1400

---

<div class="post-metadata">

**Author:** ![Bruce\_Gregory6](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/b/edb3f5/32.png) [@Bruce\_Gregory6](http://discourse.iapct.org/u/Bruce_Gregory6)\
**Post date:** [October 31, 2004, 6:45pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/8 "2004-10-31T18:45:52Z")

</div>

[From Bruce Gregory (2004.1031.1343)]

> Rick Marken (2004.10.31.0820)
> 
> Well, you just go ahead and settle back in your armchair.

If I am not mistaken, this is a fine example of a cheap shot. Giving  
Marc some competition, I see.

Bruce Gregory

---

<div class="post-metadata">

**Author:** ![Richard\_Marken2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/r/edb3f5/32.png) [@Richard\_Marken2](http://discourse.iapct.org/u/Richard_Marken2)\
**Post date:** [October 31, 2004, 8:11pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/9 "2004-10-31T20:11:04Z")

</div>

[From Rick Marken (2004.10.31.1210)]

> Martin Taylor (2004.10.31.12.04)
> 
> > Rick Marken (2004.10.31.0820)--

> > Well, you just go ahead and settle back in your armchair. I think  
> > I'll  
> > take Bill's advice and do the experiment myself if I can't find  
> > something similar in the literature.
> 
> That isn't really what I'm doing, literally or metaphorically. Quite  
> the contrary, actually. I'm trying to keep, if not my head, at least  
> my nostrils above water in the swamp of work.

I hope you didn't think I was implying that you're lazy, Martin. I know  
you're not. Far from it. What I was doing was referring to the  
appearance that you were doing "armchair science". Armchair science is  
done by people are prefer to evaluate theoretical questions by  
discussing them rather than by testing them empirically. I tried to  
suggest an empirical test of the ideas you were discussing. You  
apparently didn't think that my test was appropriate. But rather than  
suggest an alternative you just left it at that: my test was not  
appropriate. This struck me as the attitude of an armchair scientist.  
An empirical scientist -- one who does science by testing the behavior  
of models against observation of the behavior of the real thing --  
would have come up with an improved, or more appropriate, version of  
the experiment.

There's nothing wrong with being an armchair scientist. I was just  
hoping to see if we could come up with an empirical test of an aspect  
of the PCT model right here on CSGNet, rather than just always  
\_talking\_ about the model. Perhaps you could give a go at describing  
an appropriate experiment to test you ideas about bidirectionality.

Best

Rick

> **···**
>
> ---  
> Richard S. Marken  
> marken@mindreadings.com  
> Home 310 474-0313  
> Cell 310 729-1400

---

<div class="post-metadata">

**Author:** ![Martin\_Taylor2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/m/b4bc9f/32.png) [@Martin\_Taylor2](http://discourse.iapct.org/u/Martin_Taylor2)\
**Post date:** [November 1, 2004, 2:45pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/10 "2004-11-01T14:45:16Z")

</div>

[Martin Taylor 2004.11.01.09.14]

> [From Rick Marken (2004.10.31.1210)]
> 
> > Martin Taylor (2004.10.31.12.04)
> > 
> > > Rick Marken (2004.10.31.0820)--
> 
> > > Well, you just go ahead and settle back in your armchair. I think  
> > > I'll  
> > > take Bill's advice and do the experiment myself if I can't find  
> > > something similar in the literature.
> > 
> > That isn't really what I'm doing, literally or metaphorically. Quite  
> > the contrary, actually. I'm trying to keep, if not my head, at least  
> > my nostrils above water in the swamp of work.
> 
> I hope you didn't think I was implying that you're lazy, Martin. I know  
> you're not.

Well, it did look that way. Thanks for the clarification. I know what  
you mean about armchair scientists, and I agree with it.

> I was just  
> hoping to see if we could come up with an empirical test of an aspect  
> of the PCT model right here on CSGNet, rather than just always  
> \_talking\_ about the model. Perhaps you could give a go at describing  
> an appropriate experiment to test you ideas about bidirectionality.

I started to try to do that last night, but, being Halloween, the  
goblins soon found a way to cause mischief.

The mischief was actually embedded in the prior discussion. It is the  
question: "what is the question?"

My original foray into this was to ask whether there was a way within  
strict HPCT to account for a phenomenon that had not been clearly  
described in PCT-compatible language by Bruce Gregory. I tried to put  
it in PCT language, which I can paraphrase by saying that people  
often experience the (imagined) perception of a label when confronted  
by a visual scene, or experience an (imagined) visual scene when  
presented with a label. In some manner, the output of a function that  
generates the perception of either can facilitate the perception of  
the other when the data for the other are not presented to the  
senses. My (and I think Bruce's) question was whether there was a  
connection path within strict HPCT that allows for such a  
bidirectional facilitation.

Bill P. offered a kind of story that I don't fully understand, which  
involves a higher-level control system whose input is derived from  
both the picture category-level perceiver and the label  
category-level perceiver, and for which there is error if either  
input (but not both) is null. The output of that higher-level system  
invokes in some way the imagination loop for the missing  
category-level system, thereby creating the joint perception. I hope  
I've correctly abstracted the essence of what Bill was saying, though  
I can't say I quite understand how it works.

So, if Bill's presentation is fleshed out into a workable model,  
that's the answer to the original question. An experiment would have  
to probe the workings of that model, by comparing something with the  
simulation predictions of the model.

But the experiment you suggested answered a different (and  
worthwhile) question, which was to determine whether the picture  
perception and the label perception could be said to be at different  
levels, given that each independently would be evoked by sensory  
data, assuming that HPCT provided a correct model.

When I started to try to devise an experiment (which I was basing on  
the long tradition of "priming" studies and of Stroop tests), I was  
looking at yet another question. My question assumed the correctness  
of the original statement about the bidirectional experiences, and  
posed Bill's "strict HPCT" model against a model that, while possibly  
hierarchic, did not invoke control at all, but allowed for cross  
connections in which each perceptual function's output formed part of  
the other's input. I was then going to try to work out an experiment  
that did not use brain imaging (too expensive, I think, for anyone on  
CSGnet), by using priming and Stroop to affect the timing of when a  
subject would accept that a picture had formed after a label was  
presented, or that a correct label had been perceived after a picture  
was presented.

The priming would be the prior presentation of, say "airplane" before  
showing a picture of a river or of an airplane, and asking the  
subject to press a button when a label for the picture had come to  
mind, and in the other sense showing a picture of an airplane before  
showing the word "airplane" or "river" and asking the subject to  
press when an appropriate picture had come to mind.

Stroop would be a similar kind of study, except that the "airplane"  
word would be presented along with the picture of an airplane or a  
river, instead of beforehand.

In both priming and Stroop situations, a "neutral" pattern of "XXXXX"  
or a random dot pattern would serve as a baseline condition.

The experiment might give results, but then I asked myself what those  
results would tell, in respect of the two models. What would the  
models predict the effects of the priming or Stroop presentation to  
be? And I didn't know, for either model. For the cross-connected  
perceptual functions model, one might be able to make some  
qualitative predictions, but I don't know about Bill's model. Would  
the qualitative predictions be any different?

At that point, I decided to go to bed, so I could awake on All  
Saints' Day, and trusting that the saints who post on CSGnet might be  
able to help.

Martin

---

<div class="post-metadata">

**Author:** ![Richard\_Marken2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/r/edb3f5/32.png) [@Richard\_Marken2](http://discourse.iapct.org/u/Richard_Marken2)\
**Post date:** [November 1, 2004, 5:15pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/11 "2004-11-01T17:15:36Z")

</div>

[From Rick Marken (2004.11.01.0915 PST)]

> Martin Taylor (2004.11.01.09.14) --

> The experiment might give results, but then I asked myself what those  
> results would tell, in respect of the two models. What would the  
> models predict the effects of the priming or Stroop presentation to  
> be?

It's really very simple. The model would predicts that it will take longer  
to point to the category (the word "red", say) given the sensation (red)  
than it would to point to the sensation given the category. The difference  
would probably be on the order of milliseconds. There might be a clever way  
to design the study to reveal this difference consistently.

So I think you're right about Stroop-type studies being relevant. I don't  
remember the results of such studies. I'll look them up on the web when I  
get a chance. But my prediction (based on HPCT) is that it will take longer  
to identify the category (color name) given the sensation (color) than to  
identify the sensation given the category.

I don't see the relevance of whether this is done in imagination or not. I  
imagine you'd get the same result in imagination -- ie. longer time to  
imagine the category after imagining the sensation than vice versa -- it  
would just be harder to measure it.

I'll get back to you when I find some results.

Regards

Rick

> **···**
>
> --  
> Richard S. Marken  
> MindReadings.com  
> Home: 310 474 0313  
> Cell: 310 729 1400
> 
> --------------------
> 
> This email message is for the sole use of the intended recipient(s) and  
> may contain privileged information. Any unauthorized review, use,  
> disclosure or distribution is prohibited. If you are not the intended  
> recipient, please contact the sender by reply email and destroy all copies  
> of the original message.

---

<div class="post-metadata">

**Author:** ![Richard\_Marken2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/r/edb3f5/32.png) [@Richard\_Marken2](http://discourse.iapct.org/u/Richard_Marken2)\
**Post date:** [November 1, 2004, 5:39pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/12 "2004-11-01T17:39:56Z")

</div>

[From Rick Marken (2004.11.01.0940 PST)]

> Rick Marken (2004.11.01.0915 PST)---
> 
> So I think you're right about Stroop-type studies being relevant. I don't  
> remember the results of such studies. I'll look them up on the web when I get  
> a chance. But my prediction (based on HPCT) is that it will take longer to  
> identify the category (color name) given the sensation (color) than to  
> identify the sensation given the category.

Actually, I did a quick look through the Stroop literature and I don't think  
that's not where we're going to find the relevant data. The Stroop test  
presents color words in colors that don't correspond to the word. Ie. The  
word "red" written in blue ink.

What I think we need are studies of, say, visual search where a person is  
searching for a color given the category (search for the red patch) or for  
the category given the color (search for the word that is the appropriate  
label for the color of this patch). There must be a study like that that has  
been done. Can you try to find one, Martin?

Thanks

Rick

> **···**
>
> --  
> Richard S. Marken  
> MindReadings.com  
> Home: 310 474 0313  
> Cell: 310 729 1400
> 
> --------------------
> 
> This email message is for the sole use of the intended recipient(s) and  
> may contain privileged information. Any unauthorized review, use,  
> disclosure or distribution is prohibited. If you are not the intended  
> recipient, please contact the sender by reply email and destroy all copies  
> of the original message.

---

<div class="post-metadata">

**Author:** ![Martin\_Taylor2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/m/b4bc9f/32.png) [@Martin\_Taylor2](http://discourse.iapct.org/u/Martin_Taylor2)\
**Post date:** [November 1, 2004, 5:51pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/13 "2004-11-01T17:51:09Z")

</div>

[Martin Taylor 2004.11.01.1246]

> [From Rick Marken (2004.11.01.0915 PST)]
> 
> > Martin Taylor (2004.11.01.09.14) --
> 
> > The experiment might give results, but then I asked myself what those  
> > results would tell, in respect of the two models. What would the  
> > models predict the effects of the priming or Stroop presentation to  
> > be?
> 
> It's really very simple. The model would predicts that it will take longer  
> to point to the category (the word "red", say) given the sensation (red)  
> than it would to point to the sensation given the category. The difference  
> would probably be on the order of milliseconds. There might be a clever way  
> to design the study to reveal this difference consistently.
> 
> So I think you're right about Stroop-type studies being relevant. I don't  
> remember the results of such studies. I'll look them up on the web when I  
> get a chance. But my prediction (based on HPCT) is that it will take longer  
> to identify the category (color name) given the sensation (color) than to  
> identify the sensation given the category.
> 
> I don't see the relevance of whether this is done in imagination or not. I  
> imagine you'd get the same result in imagination -- ie. longer time to  
> imagine the category after imagining the sensation than vice versa -- it  
> would just be harder to measure it.

If you don't see the relevance of whether it's done in imagination or  
not, you don't see the question being asked.

The question being asked initially was whether the fact that the  
sensory perception of label often causes the (imagination) perception  
of picture, and the sensory perception of picture often causes the  
(imagination) perception of label, is consistent with strict HPCT.

It's beyond me how you can test this without reference to the fact  
that thew question is created by the bidrectionality of the influence  
on the IMAGINED perception.

Martin

---

<div class="post-metadata">

**Author:** ![Richard\_Marken2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/r/edb3f5/32.png) [@Richard\_Marken2](http://discourse.iapct.org/u/Richard_Marken2)\
**Post date:** [November 1, 2004, 6:52pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/14 "2004-11-01T18:52:33Z")

</div>

[From Rick Marken (2004.11.01.1050)]

> Martin Taylor 2004.11.01.1246]
> 
> > Rick Marken (2004.11.01.0915 PST)--

> > I don't see the relevance of whether this is done in imagination or not. I  
> > imagine you'd get the same result in imagination -- ie. longer time to  
> > imagine the category after imagining the sensation than vice versa -- it  
> > would just be harder to measure it.
> 
> If you don't see the relevance of whether it's done in imagination or  
> not, you don't see the question being asked.
> 
> The question being asked initially was whether the fact that the  
> sensory perception of label often causes the (imagination) perception  
> of picture, and the sensory perception of picture often causes the  
> (imagination) perception of label, is consistent with strict HPCT.

Maybe I would understand this better if you would tell me how I can  
demonstrate the phenomenon to myself. Is it just that when I see the word  
"red" I sometimes imagine the color red and sometimes not? Or when I see the  
color red I sometimes imagine the word "red" and sometimes not? If so, it  
seems like a phenomenon that's pretty obviously consistent with HPCT. I  
imagine the color or the word upon seeing the word or the color when I have  
some reason (goal) for producing that perception in imagination.

Regards

Rick

> **···**
>
> --  
> Richard S. Marken  
> MindReadings.com  
> Home: 310 474 0313  
> Cell: 310 729 1400
> 
> --------------------
> 
> This email message is for the sole use of the intended recipient(s) and  
> may contain privileged information. Any unauthorized review, use,  
> disclosure or distribution is prohibited. If you are not the intended  
> recipient, please contact the sender by reply email and destroy all copies  
> of the original message.

---

<div class="post-metadata">

**Author:** ![Bill\_Powers1](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/b/8e7dd6/32.png) [@Bill\_Powers1](http://discourse.iapct.org/u/Bill_Powers1)\
**Post date:** [November 1, 2004, 7:38pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/15 "2004-11-01T19:38:30Z")

</div>

[From Bill Powers (2004.11.01.1210 MST)]

Rick Marken (2004.11.01.1050) –

> Is it just that when I see the  
> word “red” I sometimes imagine the color red and sometimes not?  
> Or when I see the color red I sometimes imagine the word “red”  
> and sometimes not? If so, it seems like a phenomenon that’s pretty  
> obviously consistent with HPCT. I imagine the color or the word upon  
> seeing the word or the color when I have some reason (goal) for producing  
> that perception in imagination.

This can be treated as an automatic response (I’m pretty sure that’s how  
the idea of “association” is thought of generally), or as a  
purposive process. If it’s automatic and not purposive, then we get a  
picture of A evoking B or B evoking A without any particular goal driving  
the process. Like “content-addressed memory.” But if it’s  
purposive, we have to ask what the purpose would be – why do we want  
names to evoke images or images to evoke names? Is the purpose the same  
in either case? I doubt it. When I look at all the hundreds of things  
around me, I know I _could_ come up with at least one name for most  
of them, but I don’t unless I have some need for the name, as when  
constructing a description, or remembering the name of the person I’m  
talking to. Given names, I generally follow along in imagination with the  
thing named, because this is what I call “understanding what’s being  
said,” but unless I’m hearing directions for going somewhere or  
something critical like that, I don’t construct any elaborate or detailed  
images to go with the words – just enough so I get a sense of knowing  
what’s being talked or written about. Not that this is a vividly  
conscious process. It doesn’t get vivid unless there’s some pressing  
reason to get it right.

So, just that one comment: we have to be aware of whether we’re looking  
for a purposive or nonpurposive explanation.

Best,

Bill P.

---

<div class="post-metadata">

**Author:** ![Richard\_Marken2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/r/edb3f5/32.png) [@Richard\_Marken2](http://discourse.iapct.org/u/Richard_Marken2)\
**Post date:** [November 1, 2004, 10:27pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/16 "2004-11-01T22:27:16Z")

</div>

[From Rick Marken (2004.11.01.1430)]

> Bill Powers (2004.11.01.1210 MST)]
> 
> Rick Marken (2004.11.01.1050) --
> 
> > Is it just that when I see the word "red" I sometimes imagine the color red  
> > and sometimes not? Or when I see the color red I sometimes imagine the word  
> > "red" and sometimes not?
> 
> This can be treated as an automatic response (I'm pretty sure that's how the  
> idea of "association" is thought of generally), or as a purposive process.

Good point. I guess my point was that I know the kind of imaginative  
associations Martin described only as part of a purposive process.

Best

Rick

> **···**
>
> --  
> Richard S. Marken  
> MindReadings.com  
> Home: 310 474 0313  
> Cell: 310 729 1400
> 
> --------------------
> 
> This email message is for the sole use of the intended recipient(s) and  
> may contain privileged information. Any unauthorized review, use,  
> disclosure or distribution is prohibited. If you are not the intended  
> recipient, please contact the sender by reply email and destroy all copies  
> of the original message.

---

<div class="post-metadata">

**Author:** ![Bruce\_Gregory6](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/b/edb3f5/32.png) [@Bruce\_Gregory6](http://discourse.iapct.org/u/Bruce_Gregory6)\
**Post date:** [November 1, 2004, 10:36pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/17 "2004-11-01T22:36:22Z")

</div>

[From Bruce Gregory (2004.1101.1736)]

> Rick Marken (2004.11.01.1430)
> 
> Good point. I guess my point was that I know the kind of imaginative  
> associations Martin described only as part of a purposive process.

This assumption can only be confirmed by applying the Test, according  
to my limited understanding of PCT.

Bruce gregory

---

<div class="post-metadata">

**Author:** ![Richard\_Marken2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/r/edb3f5/32.png) [@Richard\_Marken2](http://discourse.iapct.org/u/Richard_Marken2)\
**Post date:** [November 2, 2004, 4:58am UTC](http://discourse.iapct.org/t/attaching-labels/1349/18 "2004-11-02T04:58:25Z")

</div>

[From Rick Marken (2004.11.01.2100)]

> Bruce Gregory (2004.1101.1736)--
> 
> > Rick Marken (2004.11.01.1430)
> > 
> > Good point. I guess my point was that I know the kind of imaginative  
> > associations Martin described only as part of a purposive process.
> 
> This assumption can only be confirmed by applying the Test, according  
> to my limited understanding of PCT.

Yes. That's formally true. My own experience, however, suggests that  
it's very likely that thinking the word "radio" when I look at the box  
near my computer, for example, is not something evoked by seeing the  
radio. I look over there all the time and rarely say "radio" to myself  
when I see the radio; I just use the thing. When someone asks me what  
that box is (or when I want to make a point like this on the net) I  
think of the word "radio" when I see the radio. I do this only for some  
purpose (ie. to achieve some goal) like answering a question,  
demonstrating a point, etc. But most of the time the radio is just an  
unnamed visual perception that is also the source of some nice auditory  
perceptions (sounds like a Beethoven trio). I don't think the sight of  
the radio ever just evokes the word "radio" in my mind. Whenever I  
look over there and say "radio" to myself I can always tell that there  
was some reason (purpose) for this lurking in the background.

This is just anecdotal evidence, of course. I think it would be  
interesting to see of there is any way to get evidence that an  
association \_can\_ be evoked, with no purpose behind the evocation.  
Something like the Test would have to be involved, I imagine.

RSM

> **···**
>
> ---  
> Richard S. Marken  
> marken@mindreadings.com  
> Home 310 474-0313  
> Cell 310 729-1400

---

<div class="post-metadata">

**Author:** ![Bruce\_Gregory6](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/b/edb3f5/32.png) [@Bruce\_Gregory6](http://discourse.iapct.org/u/Bruce_Gregory6)\
**Post date:** [November 2, 2004, 11:27am UTC](http://discourse.iapct.org/t/attaching-labels/1349/19 "2004-11-02T11:27:29Z")

</div>

[From Bruce Gregory (2004.1102.0627)]

> Rick Marken (2004.11.01.2100)
> 
> This is just anecdotal evidence, of course. I think it would be  
> interesting to see of there is any way to get evidence that an  
> association \_can\_ be evoked, with no purpose behind the evocation.  
> Something like the Test would have to be involved, I imagine.

I suspect that even when you are not consciously thinking the word  
"radio" you "know" the object is a radio. If I asked you, "Would you  
turn on the radio?" you would know what I wanted without having to  
figure it out. It seems to me that we know the names of objects in our  
immediate surroundings without having a purpose for this knowledge. Of  
course, I may be wrong in this regard.

Bruce Gregory

---

<div class="post-metadata">

**Author:** ![Bill\_Powers1](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/b/8e7dd6/32.png) [@Bill\_Powers1](http://discourse.iapct.org/u/Bill_Powers1)\
**Post date:** [November 2, 2004, 1:43pm UTC](http://discourse.iapct.org/t/attaching-labels/1349/20 "2004-11-02T13:43:41Z")

</div>

[From Bill Powers (2004.11.02.0623 MSR)]

Bruce Gregory (2004.1102.0627)--

> I suspect that even when you are not consciously thinking the word  
> "radio" you "know" the object is a radio.

I suspect the same thing; the "knowledge" is there in the fact that you are  
organized to retrieve the word if you have some purpose for doing so, as  
Rick said. But do you believe that whenever you look at any object, see any  
action, observe any attribute, a little whisper (or shout) of the  
auditorily-perceived or visually-perceived name appears in your brain as a  
perceptual signal?

> If I asked you, "Would you turn on the radio?" you would know what I  
> wanted without having to figure it out.

The retrieval mechanism works very fast, when needed -- at the speed of  
thought, as they say. If you point to the box and ask "What's that?" I can  
supply the name almost before the words are out of your mouth -- but there  
is a lag, because I hadn't been thinking of the name.

I'm very leery of theories that rely on "facts" that can't be observed. I  
am not aware of the names of most of the objects I'm looking at, or most of  
the other experiences I have, until there's some reason to need the name.  
My experience is like Rick's, and you have not actually said that your  
experience is any different. All you's said is that you "suspect" that you  
"know" the name of the object, meaning that your only evidence is the same  
as ours. You aren't aware of the name of the object until the name appears  
in consciousness.

> It seems to me that we know the names of objects in our  
> immediate surroundings without having a purpose for this knowledge. Of  
> course, I may be wrong in this regard.

No, you are simply using the word "know" in a fuzzy way. You mean, I think,  
that you could quickly come up with the name if there were some reason to  
do so, such as being asked for it. This says nothing about the form taken  
by this knowledge. If you take the "association" idea literally, then  
everything you see for which you "know" a word will evoke the word every  
time you see it -- you would live in a world in which everything  
incessantly and annoyingly kept telling you in words what it is. The din  
would be terrible. I see no evidence at all for that sort of explanation of  
how words are "attached" to other experiences. I don't think they're  
attached at all -- that implies an automatic process and I don't think it's  
automatic. I think we attach them, as needed. I see nothing to indicate  
otherwise.

Best,

Bill P.

[Next page](http://discourse.iapct.org/t/attaching-labels/1349.md?page=2)