The concept of the extended mind deals with external elements which, it holds, can be part of cognitive structures themselves; it suggests considering selected ones as part of the human mind, (figuratively) shattering the boundaries of skin and giving external elements the same/equal status as internal elements of the mind.
In order to be considered part of the mind, an outside object has to produce effects/results that are sufficiently comparable to those of components of the natural (internal, biological, original, classic) mind; in essence, it's about multiple realizability/functionalism.
The following (additional) criteria are initially proposed and expected to be present in accumulation; 1) constancy (the external component has to be there reliably); 2) accessibility (sort of a natural ease of use of that component); 3) automatic endorsement (the person must trust the component as they would trust any comparable part of their natural body).
Taken together (and referred to as "parity principle"), one can ask whether a mechanism/function based on/inherent to an external component can theoretically be substituted for a mechanism/function based on the natural (internal, biological etc. [see above]) mind.
Across the essays contained in this volume, the requirements are abandoned to various degrees and extent. Some abandon them rather completely (e.g. embracing a "complementarity principle") and/or take a suspiciously tangential course. Others are particularly restrictive.
Along the way, one gets to take a look at musings on the roles of language and culture. Unfortunately the presentation on what of these two remains external and what becomes internal is neither thorough, nor compelling (take a look at page 212`s "internalized exograms", though; Adam`s and Aizawa`s explanation of "non-derived" content didn`t make enough sense to me). That being said, the parallel to the more traditional ontogeny and epigenetics (DST - developmental systems theory, p. 177-180) is interesting. -- Albeit, it might very well be a false parallel; these already existing models are presumably quite capable of adopting and accounting for the features and dynamics which inspired people to come up with extended mind hypotheses.
- Making connections to the "theory of mind" could also be intriguing.
However, as far as I can tell, what's been done thus far is mostly akin to arguing about what the politically correct treatment of external elements is. At this point, there isn't much of heuristic value. As long as tools (being external - not necessarily corporeal - things) are examined carefully for their purpose, function, role and influence, there's probably not all that much need to label them parts of the mind itself. If they augment (and affect) cognition, it will be noticed. The mind works with tools, utilizes them. Until and insofar as it's not the other way around, I would keep the old distinction of human mind and tool intact. There's only need for a drastic reconceptualization once emergent properties arise. The system of (classic mind plus external element X) hasn't been shown to entail such properties of its own. It looks like its properties can be explained by the contributions (including interdependencies) of its parts. (The classic mind does have its own emergent properties and keeps them.) Also, it appears that no conceptualization of the extended mind has an equivalent to the classic mind's very own binding problem; that still exclusively plays out within the classic mind and is thus restricted to the inside.
Therefore, I - for the moment - propose thinking of a person utilizing a cane as "person utilizing cane" rather than as some kind of "unitary person-cane walking-entity" or "extended locomotioner", if you will. Nonetheless, the general idea (the extended mind hypothesis) is worth keeping in mind (excuse the pun). Under some circumstances it may actually be profitably applicable.
Lastly, I would like to recommend Mark Rowlands's text "Consciousness, Broadly Construed" (p. 271 - 294). His take on a concept of an extended mind is focused on intentionality. I personally found it to be an attractive line of thought; Terrence W. Deacon's "The Symbolic Species - The Co-Evolution of Language and the Brain" is relevant, too. - Criticism is welcome.