Charles Peirce's concept of abduction is part of the abduction/deduction dichotomy that is used for analysis when we have moved from Peirce's concept of firstness to the concept of secondness.
Abduction comes in two fundamental forms, reactive abduction and proactive abduction.
Reactive abduction is where we recognise a particular pattern as the manifestation of characteristics of a general pattern/law/principle/hypothesis. For example, seeing something fall to the ground can elicit the conscious recognition of the principles of gravity at work; we know what is behind the observed process.
Proactive abduction is where we notice what appears to be a particular (local) pattern and try to place that pattern in the 'correct' context, we seek a general pattern that incorporates the particular where initially nothing comes to mind other than the assumption that there may be meaning here.
In proactive abduction we actively place the particular in various existing or imagined contexts to try and get a 'match', to establish clear explicit meaning to the local pattern that has elicited in us an implicit sense of there being meaning in the pattern; we assume there is something behind the pattern.
What is noteworthy about abduction and deduction is that they BOTH ASSUME MEANING. For deduction we move from a given general, in the form of a hypothesis/principle/law etc, to a particular and this process adds to, or reduces, value in the general.
For abduction we move from a particular to a general and in doing so also validate the general (in the case of proactive abduction we can validate an illusion over a long period of time where, once the illusion has been asserted as 'meaningful' so it becomes part of the deduction/reactive abduction pool of contexts).
The fundamental difference between abduction/deduction and induction is that induction does not assume meaning to exist, there is no initial distinction until the inductive process moves from the particular to the general where the general is the formation of an hypothesis/law/principle. Peirce tries to explain this process in his group of signs all placed within a firstness context - Qualisign, SinSign, LegiSign. Induction thus reflects the (a) pure experience of a local followed by (b) description of one local using another as analogy and (c) the realisation of a general principle/laws behind all of these locals e.g. the experience and analysis of red, blue, green etc leads to the development of an hypothesis about principles/laws of colour. This hypothesis then becomes the context for validation; we no longer see colour 'as is' but falling within a context that sets meaning.
In the move to secondness, where a 'meaningful' context is assumed to implicitly or explicitly exist, so the unilateral movement (particular to general) found in firstness becomes bilateral (abduction - particular to general, deduction - general to particular), this process emphasises the use of dichotomisations in analysis. The group of signs Peirce related to a context of secondness is Icon, Indices, Symbol.
In the context of primary and secondary thinking processes (discussed elsewhere on this website), the abduction/deduction loop is an example of secondary processing where relational considerations takeover from the initial identification process (that has succeeded in reactive abduction and initially failed in proactive abduction other than the primary identification of the pattern); abduction/deduction either explicitly or implicitly assume a general context within which local distinctions are made and as such are resistant to random processes where there is nothing 'behind' an observed pattern (see article on markov chains).
Classical Science was more firstness oriented where observation was paramount and we did not move past the identification process. Modern Science, with its emphasis on refutability (in principle) and so a recognition of negation combined with an attraction to discover what is BEHIND what has been identified, favours secondness/thirdness as fundamental to establishing 'meaning'; firstness, at the base level, takes an experience 'as is' and does not initially make any assumptions re meaning etc. This does not allow for refutability, something that can only happen when you have moved into secondness/thirdness processing.
We can see the recognition of this in Karl Popper's approach to Science where refutability is paramount in 'doing' Science.
Abduction can be seen as reflecting inductive processes but within a context, either existing or ASSUMED TO EXIST. We can identify creativity processes at work here in that as we attempt to link a local pattern to a general (and so include negation in our analysis) so the placing of that pattern in real or imagined contexts can lead to 'innovations' in that an unexpected link is made.
Adaptive creativity is where, given a context, we vary the local pattern and so allow for variations on a theme. In this sense we hold the context as invarient and vary the text (local pattern) as compared to the innovative creative process of making the text (local pattern) invarient and vary the context.
<To come .. moving to thirdness ..)