seeker wrote:Are you trying to use empirical data to account for an inferred construct? If so, that's weird. Inferred constructs are proposed to account for empirical data, not the other way around.

I think the presumed cause of something should be testable at least to some extent independently of it's presumed effects, since something cannot be at the same time cause and effect of itself. But for cognitive processes at the moment there's no way around it besides controversial neurological data.

seeker wrote:It's not just a "metaphor", it's a functional relation between physiological and behavioral classes of events, and it qualifies as a scientific theory, IMO.

For that I think it'd have to move from correlation to causation.

seeker wrote:We were talking about cognitive theories, not about psychoanalytic speculations. Are you arguing that they're similar regarding these issues? If so, you'll need to offer concrete examples of this kind of problem in cognitive theories.

Yeah, because even while cognitive psychology accepts methodological behaviorism and psychoanalysis doesn't, and that translates into precise measurements and experiments, in both cases inferred constructs go beyond the data, whilst behaviorist models (using intervening variables "in the strict sense") are the data, you can't accept one and reject the other, as suggested by Meehl.
For instance: "There is a broad consensus among many researchers that schema theories are unprincipled. This stems from the fact that it is often possible to create any particular content for the knowledge structures used, to account for the pattern of evidence found. Schank deals, in part, with this problem by attempting to delimit all the possible structures in long-term memory, but the theory is still underspecified. Problems still remain; for example, what are the specific contents of all of these structures? In general, then, schema theories tend to be good at accounting for results in an ad hoc fashion..." (taken from Eysenck and Keane, 'Cognitive Psychology', 2000).
Bottom line: Useful and interesting models? Certainly. Scientific theories? No (at least for now).

seeker wrote:http://plato.stanford.edu/entries/constructive-empiricism/#1.5

Thanks!

Asta666 wrote:

seeker wrote:Are you trying to use empirical data to account for an inferred construct? If so, that's weird. Inferred constructs are proposed to account for empirical data, not the other way around.

I think the presumed cause of something should be testable at least to some extent independently of it's presumed effects, since something cannot be at the same time cause and effect of itself. But for cognitive processes at the moment there's no way around it besides controversial neurological data.

We agree that testability is necessary, but I don't understand why do you think that cognitive processes are not testable. We might follow Wilson's classification of theoretical constructs (http://www.ijpsy.com/volumen1/num2/23/s ... pes-EN.pdf) in:
* Abstractive Constructs (directly observable in current field of observation, e.g. stimuli, responses, reinforcement)
* Type I Hypothetical Constructs (directly observable, but in some field not currently being observed, e.g. memory traces as neural trace in cognitive psychology)
* Type II Hypothetical Constructs (in-principle unobservable, e.g. Id, ego, Piaget's schema)
* Type III Hypothetical Constructs (in-principle observable within current field of observation, but not observable for some technical reason, e.g. black holes, sometimes reinforcement histories are appealed to, which have not in fact been directly observed, but are in-principle observable).
Given this classification, you are arguing against a Realism about Type II Hypothetical Constructs (with the same argument that Wilson proposed in the linked article), and I agree with this anti-realist argument. But (1) cognitive theories don't need to include Type II Hypothetical Constructs, and (2) cognitive theories that include Type II Hypothetical Constructs don't need to adopt a realist stance toward them. So your argument doesn't apply to the cognitive theories that use hypothetical constructs of types I and III, nor to the cognitive theories that adopt an instrumentalist stance towards hypothetical constructs of type II.

Asta666 wrote:

seeker wrote:It's not just a "metaphor", it's a functional relation between physiological and behavioral classes of events, and it qualifies as a scientific theory, IMO.

For that I think it'd have to move from correlation to causation.

Why do you think that the empirical relation needs to be of a causal kind in order to qualify as a scientific theory? Many areas of research are mostly taxonomic (e.g. broad areas of chemistry, zoology, or botanics) or mostly observational (e.g. sociology, astronomy, paleontology). Would you consider them unscientific for that reason?

Asta666 wrote:

seeker wrote:We were talking about cognitive theories, not about psychoanalytic speculations. Are you arguing that they're similar regarding these issues? If so, you'll need to offer concrete examples of this kind of problem in cognitive theories.

Yeah, because even while cognitive psychology accepts methodological behaviorism and psychoanalysis doesn't, and that translates into precise measurements and experiments, in both cases inferred constructs go beyond the data, whilst behaviorist models (using intervening variables "in the strict sense") are the data, you can't accept one and reject the other, as suggested by Meehl.
For instance: "There is a broad consensus among many researchers that schema theories are unprincipled. This stems from the fact that it is often possible to create any particular content for the knowledge structures used, to account for the pattern of evidence found. Schank deals, in part, with this problem by attempting to delimit all the possible structures in long-term memory, but the theory is still underspecified. Problems still remain; for example, what are the specific contents of all of these structures? In general, then, schema theories tend to be good at accounting for results in an ad hoc fashion..." (taken from Eysenck and Keane, 'Cognitive Psychology', 2000).

Schema theories include Type II Hypothetical Constructs (following Wilson's classification), but not all cognitive theories include constructs that fall in this category. Your argument won't apply to cognitive constructs of types I and III, nor to instrumentalist stances toward cognitive constructs of types II. Also, what you've called "behaviorist models" often include hypothetical constructs of types I and III.

Asta666 wrote:Bottom line: Useful and interesting models? Certainly. Scientific theories? No (at least for now).

How do you define "scientific theory"?

seeker wrote:Given this classification, you are arguing against a Realism about Type II Hypothetical Constructs (with the same argument that Wilson proposed in the linked article), and I agree with this anti-realist argument. But (1) cognitive theories don't need to include Type II Hypothetical Constructs, and (2) cognitive theories that include Type II Hypothetical Constructs don't need to adopt a realist stance toward them. So your argument doesn't apply to the cognitive theories that use hypothetical constructs of types I and III, nor to the cognitive theories that adopt an instrumentalist stance towards hypothetical constructs of type II.

I liked that article. Can you give me some examples of cognitive theories that only rely on type I (which would be the same as Meehl's "intervening variables in the strict sense") or III constructs? What does it mean "not to adopt a realist stance towards them"?
An instrumentalist stance towards hypothetical constructs would mean something like to "to the extent that the instrument allows to judge"? Like through the MMPI or the BigFive Questionnaires? That would be fine to me as long as we can agree on the way the instrument is built (validity and reliability). In my previous messages I also said that I agree that it's more plausible in some cognitive domains than in others and that "I'm just stating some dangers that I see in taking in some cases hypothetical constructs in Meehl's sense as causes of behavior without solid neural evidence and without being able to choose among competing models by behavioral experiments." not claiming that anything that implies constructs is equally unreliable.

seeker wrote:Why do you think that the empirical relation needs to be of a causal kind in order to qualify as a scientific theory? Many areas of research are mostly taxonomic (e.g. broad areas of chemistry, zoology, or botanics) or mostly observational (e.g. sociology, astronomy, paleontology). Would you consider them unscientific for that reason?

Of course not, but to me a theory states cause-effect relationships. At correlational level you can't distinguish what is the cause and what the effect, and generally you also can't control strange variables to the extent of an experiment. In this case it's problematic because it's not merely a descriptive classification as in zoology, but causal efficacy of cognitive phenomena is commonly assumed.

Science = make prediction, collect data, compare prediction and data, improve model, make new predictions.
Psychology is a science (psychoanalysis doesn't deserve to be named "psychology"). It moves forward by eliminating falsities. It collects data methodically, and does the best to integrate the data in the theory.

What more do you want to call something a science ? Epistemology and philosophy of science/knowledge have failed to find definitive criteria to separate "science" from "not science". Even physicists don't know what is ontologically true. Completely different perspectives (ontologically speaking) in physics lead to the same predictions (see Feynman - The caracter of physical law). Physicists always use ultra simplified model of the reality, because we can't even fully modelise an atom of helium. So what ? Science is in the method. Modern psychologists use the scientific method => psychology is a science.

If I apply scientific method to voodoo is voodoo a science.?

If I apply scientific method to economics is it a science?

If I apply scientific method to politics - is it a science?

If I apply scientific method to astrology is it a science?

sorry - won't wash.....psych is a discipline like others that draws on math and science

I'd give engineering the nod long before psych and related disciplines.

Macdoc wrote:If I apply scientific method to voodoo is voodoo a science.?

If I apply scientific method to economics is it a science?

If I apply scientific method to politics - is it a science?

If I apply scientific method to astrology is it a science?

Yes to all of the above, in a sense. That is, "voodoo" wouldn't be a science, but the "study of voodoo" would be scientific. The problem above though is that all of your examples are applications or applied fields, except for economics (which already is a science, at least microeconomics is).

Macdoc wrote:sorry - won't wash.....psych is a discipline like others that draws on math and science

I'd give engineering the nod long before psych and related disciplines.

Engineering is an applied field though, so can never be a science. Psychology isn't an application of science and maths, but is the practice of science and maths. Are you perhaps thinking of clinical psychology?

If a field studies natural phenomena using the scientific method, using falsifiable and testable models, with repeatable and replicable results, and develops highly accurate quantifiable theories to predict future instances, then what more is there to science? What else does psychology need to do?

Mr.Samsa wrote:I can't find any reputable resources at the moment that give any definitive statements on how these terms are, or should be, used - are you aware of any?

No, I don't know any reputable resources. And, given such lack of reputable resources, I wouldn't be as asseverative as you've been. For example, you seem to adopt a strong distinction between "science" and "applied fields" or "mathematics". I think it's a reasonable stance, but it seems to be one semantic convention among others. If you claim that your semantic convention is more accurate than others, you'll need some support for that claim. Otherwise, you can adopt a less asseverative approach by offering your own semantic convention as an adequate one, and asking your interlocutor to give evidence that his semantic convention is at least as adequate as yours.

seeker wrote:

Mr.Samsa wrote:I can't find any reputable resources at the moment that give any definitive statements on how these terms are, or should be, used - are you aware of any?

No, I don't know any reputable resources. And, given such lack of reputable resources, I wouldn't be as asseverative as you've been. For example, you seem to adopt a strong distinction between "science" and "applied fields" or "mathematics". I think it's a reasonable stance, but it seems to be one semantic convention among others. If you claim that your semantic convention is more accurate than others, you'll need some support for that claim. Otherwise, you can adopt a less asseverative approach by offering your own semantic convention as an adequate one, and asking your interlocutor to give evidence that his semantic convention is at least as adequate as yours.

I can't answer for Mr.Samsa, but given that the separation between science, applied science, and mathematics is such a commonly held convention, surely there is no need for explicit evidence. I don't think I've ever come across an engineer or mathematician claiming their field was science - how could they when they don't utilise the scientific method?
This in no way implies that they are inferior to the sciences, but only points out that the applied fields/mathematics concern themselves with ultimate goals which are separate to those of the sciences, and thus have no need for the scientific method.

Macdoc wrote:If I apply scientific method to voodoo is voodoo a science.?

If I apply scientific method to economics is it a science?

If I apply scientific method to politics - is it a science?

If I apply scientific method to astrology is it a science?

sorry - won't wash.....psych is a discipline like others that draws on math and science

I'd give engineering the nod long before psych and related disciplines.

You just can't apply scientific method to politics nor religion. Maybe the criteria is to be "a knowledge oriented discipline". Well, "making refutable claims" seems a good criteria.
Astrology makes refutable claims. If they followed the method => they would see it's bullshit and the discipline should collapse.
Microeconomics is a science, it's just not very fruitful (and economists seem to have a bad scientific education).
Sociology and history would bother me more. Those are usefull as descriptive disciplines, but when they try to predict, they are generally bullshiting. It's just impossible to interpret the data, and "societies" can't be tested in controlled experiments. So... not sciences to me, because we can't apply the good old scientific method, but it's probably arguable for hours.

I totally agree with Mr.Samsa on applied fields.

The problem in this discussion, as ever, is a terminology problem. What do we call psychology ? We don't have clear-cut definitions and categories. And we will never have. Clinical psychology isn't a science, it's an application (and it does change everything, and there is probably a clear and definitive way to separate sciences from applied sciences). Psychoanalysis is bullshit.
Cognitive psychology is plainly a science. We refine our models according to carefully obtained data, and make predictions. We learn epistemology, statistics. Our models simplify reality ? Physicist models too. We are peer-reviewed, publish in scientific journals, and we have already understood great things. It's one of the most interesting and growing science at the moment. But, well, some people will want to call that "cognitive neuroscience". Never ending debate.
Same for neuropsychology. It's more unique case-oriented. Double dissociations (one patient, after a lesion, can do X but not Y. Another patient can do Y but not X) are very strong arguments for the modularity of the brain, and neuropsychology has produced extremely usefull models to comprehend the human psychology. Those models, based on lesions study and exceptionnally gifted individuals give limits on the potential of what a brain can do. We can predict very much on what will happen on the behaviour and the possibilities of rehabilitation given the localisation of a cerebral lesion : that's a great achievement. It's a science, not a single doubt on it.

Social psychology, personality tests, and so, it might be more debattable. To take my case, I don't know a damn thing (well, in fact I do a little, but I could know nothing) about those, although I would define myself as a psychologist. They are totally different disciplines than "cognitive psychology".
The problem may be that when we say "psychology", people don't think "scientific psychologists". And they are probably right. Psychology has a collection of different connotations. So I don't know if "Psychology" is a real science. But you just CAN'T argue against cognitive and neuropsychology.

AndreD wrote:

seeker wrote:

Mr.Samsa wrote:I can't find any reputable resources at the moment that give any definitive statements on how these terms are, or should be, used - are you aware of any?

No, I don't know any reputable resources. And, given such lack of reputable resources, I wouldn't be as asseverative as you've been. For example, you seem to adopt a strong distinction between "science" and "applied fields" or "mathematics". I think it's a reasonable stance, but it seems to be one semantic convention among others. If you claim that your semantic convention is more accurate than others, you'll need some support for that claim. Otherwise, you can adopt a less asseverative approach by offering your own semantic convention as an adequate one, and asking your interlocutor to give evidence that his semantic convention is at least as adequate as yours.

I can't answer for Mr.Samsa, but given that the separation between science, applied science, and mathematics is such a commonly held convention, surely there is no need for explicit evidence. I don't think I've ever come across an engineer or mathematician claiming their field was science - how could they when they don't utilise the scientific method?
This in no way implies that they are inferior to the sciences, but only points out that the applied fields/mathematics concern themselves with ultimate goals which are separate to those of the sciences, and thus have no need for the scientific method.

It's also a commonly held convention to distinguish between "basic and applied sciences" and between "empirical and formal sciences", so I think you need some evidence if you want to claim that one convention is more adequate than another.

Asta666 wrote:

seeker wrote:Given this classification, you are arguing against a Realism about Type II Hypothetical Constructs (with the same argument that Wilson proposed in the linked article), and I agree with this anti-realist argument. But (1) cognitive theories don't need to include Type II Hypothetical Constructs, and (2) cognitive theories that include Type II Hypothetical Constructs don't need to adopt a realist stance toward them. So your argument doesn't apply to the cognitive theories that use hypothetical constructs of types I and III, nor to the cognitive theories that adopt an instrumentalist stance towards hypothetical constructs of type II.

I liked that article. Can you give me some examples of cognitive theories that only rely on type I (which would be the same as Meehl's "intervening variables in the strict sense") or III constructs? What does it mean "not to adopt a realist stance towards them"?

I don't think type I HC are equivalent to Meehl's intervening variables (which "merely abstract the empirical relationships"). They are Meehl's hypothetical constructs ("entities or processes not among the observed"), but their non-observability is constrained to a specific context.
"Not to adopt a realist stance" means adopting an instrumentalist stance, it's saying "I don't assume that X exists, I just claim that our empirical data occurs as if X exists". For example, if you think electrons exist, you're a realist about electrons, and if you are agnostic about their existence but you claim that our empirical data occurs as if electrons exist, you're an instrumentalist about electrons.
I think most cognitive theories only rely on type I and tipe III constructs. Most cognitive theorists think that they're proposing abstract models of events that occur in the behavioral and physiological level (some of which are observable now, and some of which might be observable in the future). Connectionism, enactivism, and biobehavioral analysis are research programs that explicitly try to specify the relations between cognitive abstract models and events that occur in the behavioral and physiological level.

Asta666 wrote:An instrumentalist stance towards hypothetical constructs would mean something like to "to the extent that the instrument allows to judge"? Like through the MMPI or the BigFive Questionnaires?

No. You're talking about an operationalist stance. An instrumentalist stance is saying "I don't assume that X exists, I just claim that our empirical data occurs as if X exists".

Asta666 wrote:That would be fine to me as long as we can agree on the way the instrument is built (validity and reliability). In my previous messages I also said that I agree that it's more plausible in some cognitive domains than in others and that "I'm just stating some dangers that I see in taking in some cases hypothetical constructs in Meehl's sense as causes of behavior without solid neural evidence and without being able to choose among competing models by behavioral experiments." not claiming that anything that implies constructs is equally unreliable.

Not all hypothetical constructs are assumed to be "causes". You seem to dismiss non-causal explanations, but I see no reason for that.

Asta666 wrote:

seeker wrote:Why do you think that the empirical relation needs to be of a causal kind in order to qualify as a scientific theory? Many areas of research are mostly taxonomic (e.g. broad areas of chemistry, zoology, or botanics) or mostly observational (e.g. sociology, astronomy, paleontology). Would you consider them unscientific for that reason?

Of course not, but to me a theory states cause-effect relationships.

Would you deny the status of "theory" to a set of models that don't state cause-effect relations, but describe other kinds of relations? Why? I think many scientific theories don't describe cause-effect relations (e.g. there're taxonomical theories in chemistry, zoology, or botanics, and there're descriptive theories in sociology, astronomy, paleontology).

Asta666 wrote:At correlational level you can't distinguish what is the cause and what the effect, and generally you also can't control strange variables to the extent of an experiment. In this case it's problematic because it's not merely a descriptive classification as in zoology, but causal efficacy of cognitive phenomena is commonly assumed.

There's more than one meaning of "cognitive phenomena", so I wouldn't overgeneralize about it. Sometimes it's the explanans, and sometimes it's the explanandum. Sometimes it occurs in the behavioral level, and sometimes in the physiological level. Sometimes it's conceptualized as internal to the brain, and sometimes it's conceptualized as extended over the environment. The analysis would be different in each case.

seeker wrote:

AndreD wrote:

seeker wrote:

Mr.Samsa wrote:I can't find any reputable resources at the moment that give any definitive statements on how these terms are, or should be, used - are you aware of any?

No, I don't know any reputable resources. And, given such lack of reputable resources, I wouldn't be as asseverative as you've been. For example, you seem to adopt a strong distinction between "science" and "applied fields" or "mathematics". I think it's a reasonable stance, but it seems to be one semantic convention among others. If you claim that your semantic convention is more accurate than others, you'll need some support for that claim. Otherwise, you can adopt a less asseverative approach by offering your own semantic convention as an adequate one, and asking your interlocutor to give evidence that his semantic convention is at least as adequate as yours.

I can't answer for Mr.Samsa, but given that the separation between science, applied science, and mathematics is such a commonly held convention, surely there is no need for explicit evidence. I don't think I've ever come across an engineer or mathematician claiming their field was science - how could they when they don't utilise the scientific method?
This in no way implies that they are inferior to the sciences, but only points out that the applied fields/mathematics concern themselves with ultimate goals which are separate to those of the sciences, and thus have no need for the scientific method.

It's also a commonly held convention to distinguish between "basic and applied sciences" and between "empirical and formal sciences", so I think you need some evidence if you want to claim that one convention is more adequate than another.

I don't see how that disagrees with my post. Applied science is not 'science' in the way the word is commonly used (ie. with regard to those disciplines which utilise the scientific method), the same is true for 'empirical sciences' (normally just termed 'science') and 'formal sciences' (usually not grouped as 'science' as they only retain the word 'science' from it's pre-scientific method origin).

AndreD wrote:

seeker wrote:It's also a commonly held convention to distinguish between "basic and applied sciences" and between "empirical and formal sciences", so I think you need some evidence if you want to claim that one convention is more adequate than another.

I don't see how that disagrees with my post. Applied science is not 'science' in the way the word is commonly used (ie. with regard to those disciplines which utilise the scientific method), the same is true for 'empirical sciences' (normally just termed 'science') and 'formal sciences' (usually not grouped as 'science' as they only retain the word 'science' from it's pre-scientific method origin).

I don't think there's a strong disagreement between you and me. We can make a bridge between the two semantic conventions by saying that there's a "narrow concept of science" (e.g. empirical, basic, mainly experimental sciences) and a "broad concept of science" (which includes non-empirical, and non-basic, and non-experimental sciences). I don't think one of these concepts is "more legitimate" or "more adequate" than the other: it would depend on which features we want to include or exclude from our reference, and whether we want to refer to a narrower or a broader set of disciplines. The broader set has been mapped with several methodologies, and here you can see a review of those maps:
Klavans & Boyack. (2008). Toward a consensus map of science.
https://docs.google.com/viewer?a=v&q=cache:dCrrLR-uUzYJ:sci.slis.indiana.edu/klavans_2009_JASIST_60_455.pdf+&hl=es&gl=ar&pid=bl&srcid=ADGEESicZ1teuCnX_42RMrpb5pcJLL3xhJdHAednfP0ka978X4kboD1b99qO6Q_qpsDTyzJ21OtWvpSWhcjSvbUh7WTIwjJniuwVrnYfBluN3lT0Tqbb3zY4kSIqa5lyS5Gf_ylcXZLN&sig=AHIEtbRTZm24SZCrGcgC3uH5uvEIehJSpw

seeker wrote:

Mr.Samsa wrote:I can't find any reputable resources at the moment that give any definitive statements on how these terms are, or should be, used - are you aware of any?

No, I don't know any reputable resources. And, given such lack of reputable resources, I wouldn't be as asseverative as you've been. For example, you seem to adopt a strong distinction between "science" and "applied fields" or "mathematics". I think it's a reasonable stance, but it seems to be one semantic convention among others. If you claim that your semantic convention is more accurate than others, you'll need some support for that claim. Otherwise, you can adopt a less asseverative approach by offering your own semantic convention as an adequate one, and asking your interlocutor to give evidence that his semantic convention is at least as adequate as yours.

Well the issue over whether applied and formal sciences are science is a much easier one to answer. At the very least, we can reject engineering as a science in that including it as a science essentially opens up everything that utilises an aspect of science as being "scientific". This makes things like history a science, and politics, and even astrology or homeopathy (as they can rely on results of astronomy and chemistry, respectively). Literature and art in general would also be a science, or walking your dog. So we can reject such a distinction based purely on a desire to avoid reducing science to the absurd. In a more formal sense, we can reject it based on practically every philosopher of science's attempt to describe "science", in that a common feature among all concepts of science include empiricism and experimentation - which automatically excludes mathematics and the applied sciences, and that's even before we get to issues of falsificationism and paradigms.

As I mentioned in my comments to SeriousCat though, even if we want to take the broad understanding of 'science' which includes the formal and applied areas (which I had no problem accepting as long as it was understood to be a broader interpretation), this is irrelevant to the discussion we were having over psychology. That is, suggesting and implying that psychology needs to strive towards the scientific ideal of mathematics or some applied science is nonsensical. Under a broad definition they may all be "science", but they are clearly distinct areas with vastly different methodologies and approaches to their subject matter. So if psychology suddenly adopted the method used by mathematicians, or did their science in the way psychiatrists carry out their work, then undeniably the scientific output of psychology would drop to zero.

This makes things like history a science, and politics, and even astrology or homeopathy

I was under the impression that all behaviorists thought astrology was science. I am pretty sure if I look back far enough in the forum that Mr. Samsa said he thought so

SteveBrewer wrote:

This makes things like history a science, and politics, and even astrology or homeopathy

I was under the impression that all behaviorists thought astrology was science. I am pretty sure if I look back far enough in the forum that Mr. Samsa said he thought so

Welcome back, Steve, it's been a while.

Mr.Samsa wrote:
Welcome back, Steve, it's been a while.

Thanks, it has been awhile. I fell down the rabbit hole of trying to disentangle placebo effects from non-effects and I am in the middle of my preliminary exams at the moment. Good to see you are still around. At any rate, sorry, didn't mean to derail the discussion.

SteveBrewer wrote:Thanks, it has been awhile. I fell down the rabbit hole of trying to disentangle placebo effects from non-effects and I am in the middle of my preliminary exams at the moment. Good to see you are still around. At any rate, sorry, didn't mean to derail the discussion.

Ah yeah, getting lost in your work tends to happen when you're investigating interesting things. And no problem, it's not a major derail - hopefully you'll get a chance to contribute some more once your exams are over!

seeker wrote:I don't think type I HC are equivalent to Meehl's intervening variables

You are right, I confounded them. They'd be of the "abstractive" type in Wilson's terminology.

seeker wrote:I think most cognitive theories only rely on type I and tipe III constructs.

Type I I can imagine, could you give me some examples of type III ones?
Anyway, for type I (or as you said "sometimes in the physiological level") I agree with Wilson that: "This does not to invalidate the direct study of the relations between behavior and neural events or environmental influences on neural events. What it suggests is that speculation about these relations ought to take place in the context of direct experimental interactions with the events about which one is theorizing. That is, theorizing about neurology ought to involve direct interaction with neural events."
"Modern cognitive psychology may be used to illustrate some problems with reductive analyses. According to some cognitive psychologists “mental processes are brain processes” (Ellis and Hunt, 1983, p.11). While there may be instances where observation is not possible for technical reasons, why would one speculate about neurological events when one could examine those events directly?"
It's a common thing to read that all cognitive processes take place in the brain, but that doesn't translate into every cognitive model being well justified at neural level. For instance I've heard many times Chomsky saying UG is an inherited "mental organ" that's a function of the brain but I'm not aware there's much neural evidence to back this up.

seeker wrote:Would you deny the status of "theory" to a set of models that don't state cause-effect relations, but describe other kinds of relations? Why? I think many scientific theories don't describe cause-effect relations (e.g. there're taxonomical theories in chemistry, zoology, or botanics, and there're descriptive theories in sociology, astronomy, paleontology).

I think taxonomies assume theories (of observational instruments, statistics, etc.) but are not theories themselves because they don't explain why things are that way. For instance the DSM classifies mental disorders (and that classification assumes theories) but doesn't give any statement regarding what causes them to be that way. Anyway I don't think this is a central issue to our discussion, because causal efficacy of cognitive constructs like schemas is commonly assumed.

seeker wrote:Sometimes it occurs in the behavioral level, and sometimes in the physiological level. Sometimes it's conceptualized as internal to the brain, and sometimes it's conceptualized as extended over the environment.

That's fine but the problem is for Type I constructs there is only neural evidence so we'd be talking about neurology and not psychology. At behavioral level it could only be behavioral evidence. We are left with Type II ones, but those have a problem I have already mentioned and Wilson also notes in her article: "This validity is, however, not ontological validity, rather it is operational validity. The majority of psychological science considers increased ability to make remote predictions and to influence events a evidence of the ontological status of the theory. Loudan calls such an assumption convergent epistemological realism (Loudan, 1981). That is, to the extent that a theory produces many remote predictions, allows for the control of some events, generates a new lines of research, then to that extent we are satisfied that the key theoretical terms in the theory refer to “real” events. The problem with this position is that when one looks at the historical record, we can find instances of theories, such as the XIX century theory of the ethers that were productive in every modern sense of the word, yet we would now deny that the key terms of theories of the ethers referred to real events."
So, trying to sum up: What are cognitive processes? To me in general they are considered (ontologically) mental phenomena (Type II). The problem is that we can't experiment or provide public evidence with this kind of phenomena (besides some controversial self-reports and introspection, that in fact are also behaviors). So ways of empirically testing them come in two ways: neural evidence, but then we are not talking about mental phenomena, or behavioral evidence, but then again we are not talking about mental phenomena. Maybe you'll tell me again that other sciences use indirect measures too, for instance microscopes or telescopes. But in that cases measures are still of objects of the same field (biological or physical ones) while that's not the case for cognitive phenomena, it'd be more like using a video camera to record behavior.

Asta666 wrote:

seeker wrote:I think most cognitive theories only rely on type I and tipe III constructs.

Type I I can imagine, could you give me some examples of type III ones?

I think neural interpretations of cognitive terms might be type I, and dispositional or functionalist or temporally extended interpretations of cognitive terms might be type III.

Asta666 wrote:Anyway, for type I (or as you said "sometimes in the physiological level") I agree with Wilson that: "This does not to invalidate the direct study of the relations between behavior and neural events or environmental influences on neural events. What it suggests is that speculation about these relations ought to take place in the context of direct experimental interactions with the events about which one is theorizing. That is, theorizing about neurology ought to involve direct interaction with neural events."
"Modern cognitive psychology may be used to illustrate some problems with reductive analyses. According to some cognitive psychologists “mental processes are brain processes” (Ellis and Hunt, 1983, p.11). While there may be instances where observation is not possible for technical reasons, why would one speculate about neurological events when one could examine those events directly?"

There's no such dichotomy between "speculation" and "direct examination". Not all neural events can be observed directly, and not all our guesses are wild speculations. Think about connectionism: we can make neural neutwork models that incorporate both behavioral and physiological constraints, and explore mechanisms with an increasing biological plausibility. This seems to be a reasonable middle point between "speculation" and "direct examination".

Asta666 wrote:It's a common thing to read that all cognitive processes take place in the brain, but that doesn't translate into every cognitive model being well justified at neural level.

Not all cognitive researchers think that cognitive processes "take place in the brain". Proponents of extended cognition think that cognitive processes include brain, body, and environment, and therefore they cannot be translated to brain processes (for example, Alva Noë 's book: You are not your brain).

Asta666 wrote:For instance I've heard many times Chomsky saying UG is an inherited "mental organ" that's a function of the brain but I'm not aware there's much neural evidence to back this up.

Chomsky is not a cognitive psychologist (IMO, he's not even an empirical researcher). If he was the person you had in mind, we were talking about different disciplines.

Asta666 wrote:

seeker wrote:Would you deny the status of "theory" to a set of models that don't state cause-effect relations, but describe other kinds of relations? Why? I think many scientific theories don't describe cause-effect relations (e.g. there're taxonomical theories in chemistry, zoology, or botanics, and there're descriptive theories in sociology, astronomy, paleontology).

I think taxonomies assume theories (of observational instruments, statistics, etc.) but are not theories themselves because they don't explain why things are that way. For instance the DSM classifies mental disorders (and that classification assumes theories) but doesn't give any statement regarding what causes them to be that way.

I disagree. I don't think that the inclusion of causal relations is a necessary condition for a theory. For example, I think the periodical table is a taxonomical theory (it has the main characteristics that we attribute to theories).

Asta666 wrote:Anyway I don't think this is a central issue to our discussion, because causal efficacy of cognitive constructs like schemas is commonly assumed.

I disagree. There're many views about cognitive constructs, and not all the interpretations are causal.

Asta666 wrote:

seeker wrote:Sometimes it occurs in the behavioral level, and sometimes in the physiological level. Sometimes it's conceptualized as internal to the brain, and sometimes it's conceptualized as extended over the environment.

That's fine but the problem is for Type I constructs there is only neural evidence so we'd be talking about neurology and not psychology.

No, you won't be talking about neurology, because only some of the constructs of the theory have such characteristic. At best, you might argue that they're interdisciplinary theories that include both psychological and physiological constructs. But then, what's wrong with that?

Asta666 wrote:At behavioral level it could only be behavioral evidence. We are left with Type II ones, but those have a problem I have already mentioned and Wilson also notes in her article: "This validity is, however, not ontological validity, rather it is operational validity. The majority of psychological science considers increased ability to make remote predictions and to influence events a evidence of the ontological status of the theory. Loudan calls such an assumption convergent epistemological realism (Loudan, 1981). That is, to the extent that a theory produces many remote predictions, allows for the control of some events, generates a new lines of research, then to that extent we are satisfied that the key theoretical terms in the theory refer to “real” events. The problem with this position is that when one looks at the historical record, we can find instances of theories, such as the XIX century theory of the ethers that were productive in every modern sense of the word, yet we would now deny that the key terms of theories of the ethers referred to real events."

This argument at best supports fallibilism (i.e., the stance that what we think today might have to be changed tomorrow). It's true that ether and flogiston were dismissed, but (1) they were replaced by other constructs that are still accepted today (e.g. molecules, atoms, subatomic particles, fields), (2) we can enumerate many other examples where initially speculative constructs became accepted facts (e.g. that the Earth is round and is moving, that living beings are constituted by cells, that matter is constituted by atoms, that we live in a galaxy). If you dismiss cognitive constructs, you should apply the same criterion to the other examples that I've mentioned above.

Asta666 wrote:So, trying to sum up: What are cognitive processes? To me in general they are considered (ontologically) mental phenomena (Type II).

I disagree. There are lots of perspectives about "mental phenomena", and there's no agreement that "cognitive processes are mental phenomena". Most cognitive researchers are functionalists, and would probably defend the opposite relation (i.e. that mental phenomena are a subtype of cognitive processes). Other people would claim that mental phenomena cannot be reduced to cognitive processes, or that we should remain skeptic about this issue. Anyway, it depends on how "mental" is defined, and there's no agreement about that either.

Asta666 wrote:The problem is that we can't experiment or provide public evidence with this kind of phenomena (besides some controversial self-reports and introspection, that in fact are also behaviors).

Here it seems that you're referring to private responses (e.g. thoughts, imagery). The evidence is indirect, but that's enough to do research about them.

Asta666 wrote:So ways of empirically testing them come in two ways: neural evidence, but then we are not talking about mental phenomena, or behavioral evidence, but then again we are not talking about mental phenomena.

If by "mental phenomena" you're referring to private responses, they can be studied: they're behavior.

Asta666 wrote:Maybe you'll tell me again that other sciences use indirect measures too, for instance microscopes or telescopes. But in that cases measures are still of objects of the same field (biological or physical ones) while that's not the case for cognitive phenomena, it'd be more like using a video camera to record behavior.

Do you think that cognitive phenomena are not in the same field of other biological phenomena? Why do you think so?