12th Jan 1996

We have the ability to recognize patterns - in space, in time, Visual pattern matching, Verbal recognition, Concept recognition. Whenever we recognize something (physical or conceptual) we are identifying (putting a name to or a handle on) the thing from a lot of detailed information. It’s the same process as induction - from lots of information / properties / characteristics to one identifier. The one identifier can then be used as one of the properties for recognizing something else. (OO aggregation) This is one of the functions that STM must be performing on its contents.

Then in reverse recall and attention appears to be able to work together to be given an identifier / concept of a thing and then provide specific properties of it. Or given a property to look for, identify things/concepts that satisfy the property.

What about Generalization and discrimination, and abstraction. Abstraction is the formation of an idea apart from the concrete things or situations that are used to form the idea. Sounds just like recognition and induction. Abstract ideas so formed e.g. goodness, anger, whiteness, absent mindedness.

Generalization is the forming of general notions or statements that are derived from particular instances.

Recognition takes place only after an initial abstraction/induction has been performed. Subsequent information is then recognized as belonging to the same abstraction.

12th Jan 1996

There are more than just the five senses that produce information for the brain to process. The complete list is:



 Touch (contact & heat)





 Recall of the above

Sense         Information               Things/Events       Pleasure/pain

Sight          Words, Signs             articles, events      Colour, patterns

                   Signals, symbols

Touch        Texture, heat              Contact, burn        Tickle, hot

Hearing      Words, signals           noises, sounds     music, screech

Taste          Salt, Sour, Sweet       flavours                sweet, sour

Smell          musty, flowery           scent (airborne)    perfume, acrid

23rd Mar 1996

A feeling that I have experienced - I know there was something nice I wanted to do but I can't remember what it was. I remember the feeling about the goal but not the actual goal.

A goal is the recall of a final state recognized by a stimulus / feature pattern that would be reached by executing a sequence of responses - a habit. So before the parallel process is started you have an image of the final stimulus state and the parallel process recognizes completion and stops when that final state is recognized. Attention is attracted to the completion and you get the feeling of satisfying the goal.

I need a parallel process that takes the command to start the execution of a sequence, does it, loops if necessary and stops when complete. It also stops when told to start executing a different sequence. The sequence includes stimuli attentions and response productions. It also stops and attracts attention when the actual inputs do not match the sequence being executed.

4th April 1996

Subconscious action!

While executing  - conscious is thinking, interrupts on

  1.  inconsistent input with expected
  2.  finished task
  3.  unable to obtain required stimuli

Learned sequence relegated to parallel processing at particular concentration level.

How much of subconscious sequence saved in LTM when interrupt occurs, so it can learn from new situation - just inconsistent stimuli plus reaction? Or a little bit more to give context? How is the new learned fragment integrated into a sequence that can be executed? Does it need to be executed once more to lay it down in LTM? The most recent version replaces (due to recency factor in recall) the older one. This implies it must store all subconscious action sequences every time performed - or at least some moving windows worth of it (STM length?).

September 1996 - Page labeled AA

A change in stimulus level is an event which triggers the sensors to identify a stimulus. If the intensity of stimulus contrast-change amount is big then the stimulus will also interrupt attention else just a new stimulus will be sent to attention for processing. (What about a gradual change in stimulus, for example, lights being turned down in a restaurant or dusk at night? It will possibly go undetected by sensors. The sub-mind is always processing these arriving stimuli, comparing them with expected stimuli by executing past learnt experiences – sequences of stimuli. The sub-mind has always got some memory active and matching with stimuli. The sub-mind is actually comparing associations (stimulus 1 follows stimulus 2 [a change in stimulus]) with what is in memory. When an association does not match the expected then we have a reason to attract attention. It’s really comparing pairs of stimuli which capture the essence of the original event that triggered the stimulus 2 detection. Music is a good example of stimuli sequences – read article in CogSci 96 conference about baby surprise in recognizing syllable sequences.

In music we recognize the notes of the same tone as different because of duration of the note. But if the note drones on for too long we lose our ability to accurately judge its length and also habitualize it – no longer detect it consciously. Colour of door changes from white to cream – notice it, if from white to blue notice it. If it fades over 2 years from white to cream we don’t notice it. The mind has a built in rate of change detection – anything faster than it is detectable. Memory duration allows for detection of associate rate of change. Immediate change detected by sensory memory, short term change detected in STM. Long term change detected in LTM. Sensors can combine features detected within the very short term time window into one stimulus. STM can combine stimuli detected within its short term time window into a single concept/stimulus pattern. It is much harder to recognize longer term sequences of stimuli as a single stimulus but we can. We talk about rush hour and can recognize it when we see it.

Why recognize something? One reason is so we can disregard it, not interrupt the mind with trivia, no changing things, familiar and not important (feeling wise) so sub-mind continuously recognizing all incoming stimuli as familiar – finding patterns that match with memory and it stays subconscious until unfamiliarity is detected.

It is possible that the stimulus does not match any in memory and then it is totally new and it attracts attention. STM and LTM are being used to recognize patterns of sequences of stimuli. Recognizing sequence of effects – when it does not meet expectations, as in magician’s trick, we show surprise. We are always learning new effects in order to disregard them because most are irrelevant. But when it becomes important to know the effect sequence we can make it conscious.

13th Nov 1996 - Page labeled BB

Learning to recognize new stimuli:

   Change in stimulus – short term – attracts attention

   Record the stimulus value at the change time

   Long term – attracts attention – compared to LTM.

   Attention attracted if novel / new when compared with the appropriate memory – time span.

Change implies short term – different implies new / novel - similar implies recognition.

Different – novelty – attracts attention – conscious of it.

Similarity – recognized – not novel – no learning – ignore unless concentrating – repeating consciously – practiced enough – habit – can be done subconsciously – ignored.

Second encounter – recognize it – not novel – but novelty only wears off if habitualize its recognition – we have to consciously and successfully recognize something in order to habitualize its recognition and thus ignore it or have it become relevant.

New stimulus – attracts attention – conscious – long term record of it – novel feeling

Second time recognize it – familiar feeling (which is success for attention?)

Generalize, specialize – learnt with respect to stimuli, learn to distinguish different kinds of – which features (stimuli) are used to discriminate between and which are common therefore irrelevant. Before can do this inheritance recognition must be able to recognize aggregation of features – co-incidence of feature detection – happening in short term time frame. They get associated – order becomes irrelevant – but in long term order of such stimuli is important and plays bigger role in recognition and prediction of what to expect next. Start executing – gives expectation sequence when actual next stimulus matches expected (executed one) we get success – now can habitualize up to this success point in execution. This corresponds to habitual recognition which will attract attention when sequence is different.

So second time triggers execution – expected - if match – success – habit started. Third time encountered – the habit when started will throw away recognized stimuli in same sequence and make conscious only abnormal sequence – unexpected stimulus will become memorized.

In summary: 1st time stored, 2nd time executed, 3rd time habitualized.

If sometime A then B and sometimes A then C or A then D we recognize that there is no sequential, association in time, B and C not always? Learn no predictability to what follows A?

If sometimes X then Y, and sometimes W then Y, and sometimes T then Y what conclusion do we draw? - Y unpredictable?

Sequences that are predictable are then recognized as a single stimulus – in STM such a sequence can collapse into one item in STM because we recognize it as a unit. Baar pg 37.

Learning new habits – responses – do not have expected resulting stimulus – just an expected feeling that triggered the go/no go decision. Then when actual resulting stimulus matches executed one then have success.

Inner speech, mental imagery is inner dialogue, visual images.

As long as there is uncertainty / unpredictability we remain conscious of it.

Baar Pg37 – This is aggregation – grouping of features on a reliable basis provide units that can be re-recognized and reused as features of larger stimuli. A stimulus has internal consistency – it’s reliable, repeatable, predictable, and recognizable. A stable environment is a predictable environment – deterministic.

Do we need to have a feeling stimulus called novel and familiar stored with a stimulus to mark it as a unit which can be habitually recognized and ignored? We certainly have surprise as an emotional reaction to the feeling of novel events that violate our expectations!

Complexity is reduced – have a handle on it through learning.

Emphasize the importance difference between attending to an address and waiting for response such as the Tip-of-the-Tongue experience when attending to memory as the source of stimuli – implies we attend to recalls. This is attending to LTM. We also can attend to STM. Implies STM is a source of stimuli (recalls, thoughts) which are attracting attention initially.

[During the period Jan 1996 to July 1997 most of these notes / ideas were consolidated and added to an unfinished book.]