Mental Representations and Consciousness Autumn Term 2001
(This material is from Perner & Dienes (forthcoming) “A theory of implicit knowledge”, Oxford University Press)
Volition
"Action" is more than bodily movement:
An action is caused by an appropriate intention.
BUT: For many actions, there is no conscious intention preceding them. (But you did mean to do them.)
Searle (1983):
All actions are accompanied by an "intention-in-action" whether or not there is a "prior intention": "I am raising my arm by way of carrying out this intention".
(NB: An intention-in-action constitutes a higher-order thought.)
Problems:
How can we distinguish different types of actions?
1) Non-intentional slips from intentional action:
Driving to the wrong destination due to a lapse of attention:
It is an action, but it was not intended.
According to Searle, it must have had an appropriate intention-in-action, so why is it experienced as not intended?
2) Hypnotically suggested actions: Why do they feel involuntary? Surely they must have intentions in action?
Norman & Shallice (1986):
Supervisory Attentional System (SAS) (attention demanding, conscious control)
contention scheduling: selects according to level of activation, which is determined by trigger conditions of the schema and lateral inhibition/excitation between schemata.
SAS can bias activation values and is needed for
1) planning or decision making
2) Troubleshooting
3) learning new actions
4) technically difficult actions
5) overcoming strong pre-existing response
Baddeley (e.g. 1986) added: random number generation (5), dual task perf (1,3,5), attention focusing and switching (5?).
Perner, 1998
Two types of control (SAS and contention scheduling) relate to the implicit-explicit distinction.
To follow verbal instructions to implement new actions (3) and to plan (1), involve declarative representation of schema content "If S1 do A1" as a goal (often: as something desired).
(4) also presumably refers to strengthening difficult actions by declarative representations.
=> (1), (3) and (4) require fact-explicit representations of goals.
(Contention scheduling can use fact-implicit representations of conditions, actions, and goals.)
(2) and (5) require, in addition, meta-representation:
To inhibit the right competing action tendency one must represent it as implemented in my schemata - it has vehicle properties that need inhibiting.
(Consistently: correlation between passing Theory of Mind and passing executive tasks that require inhibiting previous actions - see Ted's lecture on autism.)
Perner (in press)
Vehicle control:
Action schema comes to control behaviour because of its vehicle properties i.e. degree of activation.
(predication implicit knowledge)
Content control:
Fact-explicit representation ("if condition C then do action A" or "do A") determines which schema comes to control behaviour (the schema with the conditions and action described by this representation).
Contention scheduling requires only vehicle control
Executive tasks require content control.
Content control-based system (the higher system) sets off action schemata, that then run according to vehicle control; the higher system may or may not monitor the results of the action schemata.
Sense of agency and intention comes from the higher system (a) triggering the lower system; (b) monitoring its successful performance; and (c) both the above forming and stabilizing the representation "I am intending this action".
Hierarchy of voluntariness:
Movement: Bodily movement not due to an action schema (e.g. slipping)
Action: Due to action schema (goal directed).
- Non-intentional: vehicle control produces an action not set by the higher system.
-Intentional: higher system sets up the lower system; when appropriate conditions are met, the schema executes the appropriate action.
- absent-minded intentional action: lower system runs unmonitored, producing actions appropriate to plan.
- fully content-controlled intentional action: Higher system triggers and continuously monitors execution of action schema, over-riding when necessary.
-without HOT: unconscious performance of executive function tasks??
-fully voluntary action, i.e. performed with HOT ("I intend to perform this action")
(and typically with 3rd order thought, making you aware of your intention "I know I am intending to perform this action").
Hypnosis. Three possible mechanisms:
1) Content control without a HOT?
a) Suggestion to forget the number "four": "1,2,3,5,6,.." - must be content control, but person claims ignorance of doing anything strange => no second order thought.
b) Spanos, Radtke, and Dubreuil (1982): highs suggested to forget certain words in any type of task given to them produced those words at a below baseline level in a word association test. Content control, because the existing associations that would be produced by vehicle control must be suppressed.
c) Sackeim, Nordlie, & Gur (1979). A high given strong motivation instructions for blindness performed significantly below chance in reporting the emotion shown in photographed faces.
d) Bertrand & Spanos (1985): Given a list of three words in each of three different categories, highs could selectively forget one word from each category. Subjects recalled in a category by category basis, and must have inhibited the to-be-forgotten word when recalling each category.
e) In general, virtually any arbitrary behaviour can be hypnotically suggested despite the fact that such behaviour might be novel to the person, and many hypnotic suggestions require the person ignore some salient aspect of the situation (e.g. analgesia suggestion) => many hypnotic responses are under content control.
Content control is presumably easier with relevant HOTs to support it, they can add to and support the controlling activation. Therefore, to engage in content control without HOTs, one must be good at content control. Prediction: Highs should be better than lows at content control. Is this true?
1) Graham & Evans (1977): Highs better than lows at random number generation (failure to replicate: Crawford et al 1993)
2) Naish (1983): Highs can bias perceptual schemata more strongly than lows. In a signal detection task (detect 300ms tones amongst 360ms tones), when subjects were told the signal would be more likely, the change in bias was higher for highs than lows
(based on verbal report: perception changed not just response tendency).
3) Dienes (1987): When hearing simultaneous lists of words, highs can select on the basis of representational content (semantic category) to a greater degree than lows, but they cannot filter according to purely sensory features any better than lows.
(cf Karlin, 1977; Dolby & Sheehan, 1977)
4) Dixon & Laurence (1992): Highs have greater strategic control over reversing the Stroop effect than lows.
5) Some evidence highs can maintain attentional focus better than lows (Das, 1964; Gur, 1974; Smyth & Lowry, 1983; Wallace & Patterson, 1984).
6) Some evidence that highs can co-ordinate the processing of multiple simultaneous information sources better than lows (Sigman, Phillips, & Clifford, 1985).
7) Some evidence that highs have faster reaction times to complex decision-making tasks than lows (Crawford, Horton, & Lamas, 1998).
8) Crawford et al (1993): Following analgesia suggestions, highs (but not lows) showed (by fMRI) an increase in blood flow to the anterior frontal cortex (associated with SAS function).
9) Graffin et al (1995): Highs compared to lows show more theta activity in frontal but not posterior areas. Frontal theta is claimed to indicate “the continuous concentration of attention”.
(But are highs just more motivated then lows to perform well in research they know is to do with hypnosis? Could that explain all the 9 points above?)
Relatedly: In an EEG study, Crawford et al. (1998b) documented increased response in the anterior frontal cortex of highs given an analgesia suggestion rather than no suggestion.
Notes. Other points to consider:
1) Cognitive suggestions like hallucinations can be produced by the content-controlled production of imagery; the lack of suitable HOTs about intending the imagery would lead the person to experience the image as a perception (cf Kihlstrom, 1979).
2) Hidden observer (HO): The HO can give pain ratings; i.e. express HOTs about being in pain? Does this not contradict the notion that suggestion is content control without HOTs? Two responses:
(a) HO relies on a different mechanism (formation of different "I"s; see below);
(b) HO cue simply acts as a cue for the strategic reversal of the suggestion (see Kirsch & Lynn, 1998, Psych Bull, 123, 100-115 for supportive discussion).
3) Like socio-cognitive perspective, our view emphasizes strategic nature of hypnotic responding
BUT we regard the lack of voluntariness as real;
4) like neo-dissociation theory, the "executive ego" (SAS) is involved in responding
BUT there is not two streams of consciousness;
3) Bowers and Woody
They say: hypnosis involves a weakening of frontal lobe function and that contention scheduling controls behaviour (hence the feeling of involuntariness).
We agree: frontal lobe function must be less efficient without the HOTs to support it
BUT hypnotic responding cannot be based simply upon contention scheduling. Bowers and Woody’s theory fails to get to grips with the highly strategic and, when necessary, content-controlled nature of hypnotic responding.
like Bowers and Woody, control is split off from consciousness, BUT the SAS is still involved.
4) Kirsch & Lynn (1997; American Journal of Clinical Hypnosis, paper available from me):
hypnotic responding is based on contention scheduling, and the generalized intention to perform action x whenever it is suggested; this "implementation intention" produces an automatic response (action x) whenever its satisfaction conditions (i.e. the suggestion) are obtained.
However, one cannot automatically perform tasks requiring content control, not even with generalized implementation intentions: One must specifically form the right fact-explicit representation and this goes beyond contention scheduling.
How is the normal ascent from fact-explicitness to attitude explicitness prevented?
According to HOT theory, HOTS are just thoughts and so their occurrence will be sensitive to the same influences as other thoughts (Rosenthal, 2001). That is, consistent with the socio-cognitive approach, there are two possibilities for why a HOT about intention might not occur:
a) Expectation: the expectation that the act will occur involuntarily prevents HOTS of intending from occurring.
(BUT subjects are often surprised by their hypnotic responses. Doesn't surprise indicate that something counter-expectational happened?
The expectation of X can cause X to happen, but the extent to which the expectation causes X need not match the amount of X expected; that is, our expectations can be effective in a way that is personally surprising.)
b) Congruence with beliefs:
A belief system about hypnosis might inhibit contradictory beliefs (like "I am intending this action") even in the absence of a specific expectation for a specific suggestion. This view is supported by Lynn et al's (1984) study on how beliefs about ability to resist suggestions determines general ability to resist suggestion.
2 HOTs without first order states.
Rosenthal (2000, 2001) : One can have a second order thought that one is in a certain first order state, without actually being in that state. It will seem to one that one is e.g. experiencing pain even though in fact one is not!
=> mistaken HOTs could produce many of the experiences brought about by hypnotic suggestion?
Such mistaken second order thoughts should be most likely to occur if one strongly expects the first order state to occur.
Kirsch (1985, 1991): Hypnotic experiences are the outcome of expectations (see also Kirsch & Lynn 1999 in American Psychologist).
Evidence:
1) General responsiveness of hypnotic subjects to demand characteristics.
2) For suggestions given with no induction, correlation between expectation of response and response = 0.53 (Braffman & Kirsch, 1999)
3) Post-induction, expectation of number of suggestions that will be experienced correlates highly with number of hypnotic suggestions experienced, r = .64 (Council et al, 1986).
4) Subjects pass more suggestions after an induction rather than without an induction only to the extent that they expect to (Braffman & Kirsch, 1999).
5) Subjects expectations experimentally enhanced: suggestion for seeing red was given, hidden red light bulb switched on. In later suggestions, no subject tested as a low! (Wickless & Kirsch, 1989)
False HOTs could also be facilitated by:
a) imagining the suggested events (e.g. in fly hallucination, imagining the fly) to produce sufficient first order information to trigger the primed second order thoughts (consistent with the frequent use of imagination in hypnosis - but it is not essential: see the tutorial Zamansky paper; Hargadon et al 1995);
b) being skilled at imaginative absorption (e.g. Roche & McConkey, 1990; Tellegan & Atkinson, 1974; Wilson & Barber, 1981; etc); this predicts hypnotizability when expectancy is controlled (Silva, 1990);
c) having vivid imagery (Sheehan, 19xx).
3. Different "I" representations
Kihlstrom (1997): subject creates an additional "Hypnotic I". Because the hypnotic I's intentions (causes of hypnotic responding) are not linked to the normal "I" the person does not experience himself as intending the actions to occur.
Possible mechanism, but complex.
Cannot be the basis of most hypnotic responding - it predicts a hidden observer, and a fully fledged hidden observer only occurs for about 5% of the population.
The hidden observer itself it likely to be a direct response to suggestions for there to be a hidden observer (Spanos, 1986), and some claims for the existence of a hidden observer may simply reflect incomplete responding ("I did hear an imaginary fly but part of me was aware it was just imagination").