Transcript Over kwaliteit

Mechanistic explanation and the
integration of insights from the
humanities and
cognitive sciences
Machiel Keestra, Institute for Interdisciplinary Studies,
AIS Conference, Tuscaloosa, Oct. 8, ‘09
http://home.medewerker.uva.nl/m.keestra
Overview:




2
Interdisciplinarity and theoretical pluralism in the
study of complex, multi-level systems
Integration as the aim of interdisciplinary research
Mechanistic explanation as an alternative integrative
technique
Aspects of mechanistic integration of insights of
hermeneutics with cognitive science
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Example: human action understanding
Two rivalling perspectives:
 Meaning of action requires involvement of socio-cultural
and historical contexts
 Action is determined by neural and biological factors and
requires bottom-up explanation
(compare simulation theory & theory theory)
3
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Theoretical pluralism in multi-level systems
Relative significance of theories
(restricted domain)
Multi-level systems: various constraints
on normal functioning stem from distinct
levels
Relative causal autonomy at each level
Different kinds of interactions between
levels
Intervention techniques differ per level
How does ‘hermeneutics’ fit in such
systems?






4
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
No strict localization & identity
of neural or mental functions
5
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Multiple realizability of mental functions?
Multiple functionality of mechanisms?
‘Early’ mechanisms are involved in
many functions
Developmental trajectories may differ
Learning does result in different
processing systems
Possibility of individual processing
routes
Pathology can lead to compensatory
mechanisms





6
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
An Integrated Model of the
Interdisciplinary Research Process
A.

Drawing on disciplinary insights.
1.
Define the problem or state the focus question
2.
Justify using an interdisciplinary approach
3.
Identify relevant disciplines
4.
Conduct a literature search
5.
Develop adequacy in each relevant discipline
6.
Analyze the problem and evaluate each insight to it

B.









Integrating insights and producing an interdisciplinary understanding
7.
Identify conflicts between insights and their sources
8.
Create or discover common ground
9.
Integrate insights
10.
Produce an interdisciplinary understanding of the problem and test it
(Source: Repko 2006, 123; Repko 2008, 142)
Disciplinary insights = insights from disciplines, subdisciplines, interdisciplines, schools of
thought etc.
7
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Common ground always via conflicting
insights?






Concepts
Theories
Methods
Assumptions
Values
Principles

(Newell; Repko; e.a.)
or perhaps ‘simply’:
 ‘conflicting’ evidence concerning (seemingly) similar
phenomena?
 unexpected factors modulate relatively stable phenomena?
 only a single phenomenon (pathology) requires explanation?
8
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Repko’s five techniques of integration:
According to the sources of conflict and after discovery of common
ground:
1.
2.
3.
4.
5.
theory expansion (addition of a factor e.g.)
re-definition (create common meaning)
extension (beyond original domain/time)
organization (relating re-defined concepts/assumptions)
transformation (cf. opposing assumptions into variables)
(Repko, 2008)
9
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Additional integrative technique:
Mechanistic explanation.
1. Decompose a phenomenon
2. Try to identify & localize components
& possibly sub-(sub-)components
3. Integrate these into a multi-level
mechanism
4. Differentiate intra-level & inter-level
relations/activities
5. Study role-functions & constitutive
relations
6. Re-compose the mechanism in light of
new results
(Bechtel, 2006; Craver, 2007; e.a.)
10
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Phenomenon-oriented integration




Mechanistic explanation aims at
explaining a specific phenomenon
Not so much causal but constitutive
explanation
The mechanism or components can
still be involved in other phenomena,
as well
It may turn out to be necessary to
add other components or levels, in
light of new properties.
11
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Research of mechanisms: intervention and
detection are distributed among disciplines



Level-specific theories, methods, techniques
For instance: bottom-up TMS intervention & behavioral results;
or top-down cogn. psychological task designs & single cell
recording; etc.
The more results converge, the more robust the evidence is
12
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Mechanistic integration of
hermeneutics and cognitive science?




Define all levels in terms of components &
relations - e.g. semantic and symbolic
networks & neural systems
Investigate intralevel relations, e.g. between
terms, symbols, contexts & neural networks
Investigate inter-level relations/modulations,
with intervention & detection techniques
Focus on long-term, dynamic interactions
that constitute the eventual mechanism: how
to explain historical contingencies that
influence neural systems?
13
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009
Thanks for your presence!
14
Machiel Keestra
- Mechanistic Integration
-
AIS conference Oct. 8, 2009