Brains working towards machine consciousness

In general, Research and Learning are treated as separate entities in the HE sector, which, if we look at definitions of the words, might seem a bit incongruous:

Research 

Research is work that involves studying something and trying to discover facts about it

If you research something, you try to discover facts about it

Learning is the process of gaining knowledge through studying.

Why do we think of them as separate and distinct activities?  Learning is often associated with teaching, of course, where an expert facilitates the learning experience.  Depending on the pedagogy, this can take many forms, and pedagogies can be related to research methods.

Instructivism involves the presentation of facts by the expert, which the learner is supposed to take on board.  Some will memorise the details, others will abstract rules from the examples given and learn to generalise, creating a deeper level of learning.  This latter group are, to some extent, using their own learning methodology, constructing knowledge from their experience.  If the learners rely on the first approach, it might be argued that there is no similarity with research; but in the latter case, the production of internal models and rules represent facts about the material being presented.  It is, indeed, a remarkably similar process to performing a desk-study, and in either case the idea of Cognitivism as a learning theory applies (though the rote learning can more appropriately be considered as a behavioural model).

Constructionism (Papert, Harel) works with the Constructivist theory of knowledge/learning being chiefly based on the idea that people learn best if they actually make things.  This is not dissimilar to the Constructive Research method.

Connectivism (Siemens, Downes), which considers how the learner makes connections, and is most applicable to knowledge domains with short half-lives (i.e. where the useful life period of knowledge is small due to rapid environmental changes, often true in technological areas).  This is possibly the least well connected to an established research method, although it has clear connections with collaborative and inter-disciplinary research techniques.

Rhizomatic learning (Cormier) is also focused on the rapidly changing knowledge domains, with an emphasis on community as curriculum.  The concept does not rely on pre-determined inputs from experts, but rather the continual re-negotiation of meaning within the community of learners (which can also include experts).  This is closely related to the progress of research through peer review and publication, each building on previous generations.  Of course, the publication and review process introduces significant temporal lag into the development of knowledge, and the limited space in journals also serves to restrict the quantity of new ideas being shared.  One might argue that these restrictions also serve to enforce a level of quality control, although sample rejections in the computer science domain (From IEEE "Computer") indicate just how poorly the process works in some cases.

So why do we treat the two areas separately?  Why not use the same techniques and technology to support both arenas?  In both cases, there is a period when the researcher/learner probably wants to keep their (lack of) knowledge to themselves for fear of a poor reaction from the community.  In both cases, when the researcher/learner feels more secure there is a benefit to sharing with a limited community in order to gain feedback and support (scaffolding supporting developmental thinking in the Zone of Proximal Development or ZPD(Vygotsky)).  And in both cases, there comes a point where dissemination is key - although traditionally for the learners this may be in the limited form of dissemination-by-exam-script.

I really can not see enough of a difference to be able to justify maintaining artificial barriers between the two.  As a number of people are starting to say "learning is what we do every day"