[Contrast Margaret Boden's two volume history of Cognitive Science

    Mind As Machine: A history of Cognitive Science (OUP, 2006)
    (unfortunately very expensive: I think I'll try to persuade her
    to make all the individual chapters available on her new web
    site.
]

The scientific study of cognition had a long history in Europe and continued
during the 20th century (including Freud, Piaget, Vygotsky, Gestalt
psychologists and many others).

In the USA (and to some extent other places) there was a period in the 20th
century during which psychology was dominated by behaviourists (e.g. Watson,
Skinner, and others) focusing mainly on relationships between observed,
measurable, stimuli and responses, and playing down the importance of internal
processes involving representation, learning, reasoning, etc.

So the cognitive revolution was *mainly* a shift of emphasis in the USA (and places
influenced by the American psychology), from a focus on relations beween stimuli and
observable behaviour, back to theories about *internal* processes, including work by
Bartlett, Lashley, Chomsky, Neisser, Miller (+ Galanter and Pribram) in a process that
began to engage with computationally expressed and tested theories in the 1960s, e.g.
Simon, Newell, John Anderson, McCarthy, Minsky and others.

It also used a level of description of internal processes that was different from
physiology-focused neural mechanisms, emphasising information *content* more than
information mechanisms (left to neurophysiologists).


[
I joined in around 1970 stimulated by an AI vision researcher Max Clowes, who
became known for (among other things) the slogan: 'Perception is controlled
hallucination', echoing older ideas of von Helmholtz, and more recently the
vision scientist Richard Gregory in the 1960s and later.

A lot of the new work was also inspired by the older Gestalt psychologists, e.g.
Kohler (on Apes) and Max Wertheimer (Productive Thinking), and in some places
also the Genetic Epistemology of Swiss developmental psychologist Jean Piaget.

One of the key publications in the USA was

    G.A. Miller, E. Galanter and K.H. Pribram
    Plans and the Structure of Behaviour, 1960

(Objects played a role of course -- but did not dominate.)

Chomsky's work on language processing and his critique of Skinner (a devastating review of
'Verbal Behaviour') also had a deep influence on many people (inclduing me). He stressed
the importance of internal (recursive) structural descriptions rather than just streams of
signals.

Independently of that, and some time before the "official" start of the revolution, The
neuroscientist/psychologist Karl Lashley wrote a deep and very influential paper in 1951
"The problem of serial order in behavior", arguing against behaviourist models of
behaviour as sequences of triggered reactions.

He showed that complex actions extended over time often had internal structural relations
that required cognitive mechanisms creating and using hierarchically structured intentions
and plans (that's a rough summary from memory!).

The psychologist Ulric Neisser published a book called 'Cognitive Psychology' in 1967. I
think some people claim he invented the label 'Cognitive Science', but I am not sure. He
thought that computational models could work only for non-affective states and processes
('cold cognition' not 'hot cognition').

That prompted Herbert Simon to argue that computational processes were also deeply
involved in motivation and emotion -- paper (re-)republished in 1967 as

    Motivational and emotional controls of cognition

(I was much influenced by that but I thought it had a serious mistake, namely
emphasising a "computer-like" single serial process, whereas actual human
affective and cognitive processes involve much parallel processing -- not like
the current neural nets, but concurrent streams controlling and being influenced
by perception (several streams), action (several streams, e.g. walking and
talking) as well as multiple more central processes [Ideas later expanded in the CogAff
project]. Compare Minsky's Society of Mind 1987, and later The Emotion Machine 2006.)

Anyhow, objects are involved in all of that, but any attempt to make objects *dominant*
(Spelke's idea???) seems to me to ignore both facts about animal cognition and the actual
history of cognitive science, which covers a very broad range of phenomena, including
perception, motivation, reasoning, planning, action control, learning, hypothesis
formation, language understanding and generation, and many more.

A common theme was growing recognition of the importance of computational models of
internal information processing. (Some people, especially critics, who did not understand
the variety of types of virtual machinery, thought this literally implied that cognition
was based on a sequential stream of instructions: a daft idea.)

As far as I know Spelke completely ignored most types of computational model. After she
gave an invited talk around 2006(?) to an EU cognitive robotics conference, she ended by
saying, to my amazement, that she could not see the relevance of her work to building
robots, or vice versa. I tried to get a chance to talk to her about this, but the queue
was too long and she had to leave. I recently watched a video talk about relevance of AI,
and all she could think of was how her kind of work could influence computational models,
not vice versa.

I have just found this:
https://en.wikipedia.org/wiki/Cognitive_revolution

    The cognitive revolution is the name for an intellectual movement in the 1950s
    that began what are known collectively as the cognitive sciences. It began in the
    modern context of greater interdisciplinary communication and research. The
    relevant areas of interchange were the combination of psychology, anthropology,
    and linguistics with approaches developed within the then-nascent fields of
    artificial intelligence, computer science, and neuroscience.

    A key idea in cognitive psychology was that by studying and developing successful
    functions in artificial intelligence and computer science, it becomes possible to
    make testable inferences about human mental processes. This has been called the
    reverse-engineering approach.

Regarding objects, a minor point:

One of the videos I used in my recent recorded talk (for IJCAI, and also in
Jerusalem?) showed a male weaver bird in the process of learning to make a nest
out of many long thin leaves.

When he uses such a leaf to form a loop does he create a new object? When he
pushes an end of the leaf through the loop and pulls it tight, forming a knot,
does he create a new object? When a loop forms, is the hole an object (an object
filled with empty space (air))?

As far as I recall, Spelke's work ignores the cognitive mechanisms required for
dealing with objects that don't yet exist but need to be created.

Apologies: I am rambling again. It's good that you are taking account of
existing literature, but be sure to use it critically.

However, if you are really interested in what babies perceive and react to you
can include a lot of phenomena Spelke and colleagues ignore, e.g. how children
born blind interact with the world, what infants do with their mouths (including
sucking nipples which is a highly non-trivial task in the case of real nipples,
and much easier with artificial nipples [Can you see why?]) and all the things
they do with body parts, e.g. putting fists and feet into their mouths, and
exploring body parts and other things with their hands.

A random selection:
    https://www.babycenter.in/thread/174721/baby-sucking-her-toes
    https://www.youtube.com/watch?v=xoSwCIfKTGE
    http://www.parents.com/baby/development/behavioral/the-sweetest-baby-milestones/

Baby apes, can almost immediately hang onto their mother's fur while she climbs
trees, and (some?) infant humans have similar capabilities to a lesser extent.

https://whyevolutionistrue.wordpress.com/2014/04/08/the-grasping-reflex-of-babies-a-vestigial-trait/

Some researchers will respond that these are examples of "non-cognitive"
reflexes. But I think cognition is full of reflexes (data-driven interrupts that
divert attention, redirect vision, re-organise action -- e.g. to prevent
falling, etc.).

The fact that such control processes are possible is an important aspect of
cognition. The theory of architecture-based motivation (i.e. not reward based
motivation) is another example.

Some people will argue that I am confusing cognition with other things. I then
challenge them to provide a *complete* specification of the relevant information
processing architecture and explain on what principles they segment cognition
from the rest. (In many cases, it's only defined by what they have learnt to
study.)
}}