Dear Terry,

Thanks for your reply. These propositions are indeed interesting, and
I’d agree that some of them would form the basis of serious PhD
research. I’d say that some of the individual propositions you put
forward are large enough claims that they’d support serious research
projects to demonstrate tem or prove or disprove subsidiary aspects.

While I think the list of propositions is interesting, I’m not sure
that it adds up to a coherent research program or a the foundation of a
comprehensive theoretical program. It does seem to me that
demonstrating, proving, or disproving many of the individual
propositions you state here would contribute to advancing design
research.

Three brief points occur to me.

First, the propositions that you put forward here are not implicit in
Jaime Henriquez’s rant. You’ve adduced interesting and useful
propositions by restating and building on some ideas from the rant. The
rant itself was neither informative nor helpful, and there really was no
PhD there. Your list of propositions offers issues that could lead to
serious PhD research. Jaime Henriquez’s rant did not.

Second, demonstrating causality is difficult. Proving causality is even
more difficult. For physical scientists to prove causality in fields
such as physics, medicine, or climate change requires massive evidence
across huge samples or populations. This leads, as you state, to
predictions that others can test. There is an immense amount of work
that must take place before we know enough about design to demonstrate
causality. It seems to me that work in such fields as experimental
economics and behavioral or psychology offer useful models. Project UMA
is working on a number of issues that should lead to better
understanding of how some aspects of design influence user behavior,
choices, and decisions. From this, one should eventually develop a range
of specific kinds of information that guide designer decisions. UMA is
research collaboration between TU Delft, Swinburne, Cambridge, and
Vienna. Prof. Paul Hekkert of Delft and Swinburne heads it with Prof.
Allan Whitfield from Swinburne’s National Institute for Design
Research as co-lead. Paul is also president of the Design and Emotion
Society. As you note, several people in D&E are pursuing different
aspects of this kind of research, seeking demonstrable causal
explanations.

Third, understanding and defining causality as a principle of
scientific explanation is genuinely difficult. The reason that Daniel
Kahnemann, and Vernon Smith shared a Nobel Prize for behavioral
economics is precisely that they – together with such figures as Amos
Tversky – found ways to test and demonstrate aspects of economic
choice in an empirical way. Many of the experiments they undertook
involve small, modest experiemts leding to micro-level discoveries. No
one has yet assemvled this kind of experiment into a broad theory linked
to general causation – and this is what you are calling for, in part,
for design theory.

The fourth has to do with Paola Trapani’s reply to you and the
pinboard she posted. The web site is interesting, but it does not
demonstrate casualty. Quite the contrary, it is associative.
Distinguishing causality from what the pinboard labels “perceived
causality” is one purpose of scientific experiment. 

Astrology is a case of perceived causality. Because a certain
configuration of stars was visible when a kingdom fell or a great
composer was born led to the notion that such a configuration of star
leads to the fall of a kingdom or the birth of a great composer. That is
perceived causality: there is no demonstration of causal effects, but
merely the association between two phenomena. If Paola can demonstrate
causality rather than perception here, I’ll be curious – but I
don’t see how she can do so. Political propaganda, advertising, and
astrology all rely on mistaken belief induced by perceived causality.
This is not causality, but rather persuading people to buy soap, a car,
or a political candidate through associations linked to the product or
person being sold.

Perceived causality reminds me of a conversation between two dogs where
one dog explains to the other that a well-trained scientist feeds him
whenever a bell rings. Labeling perceived causality as
“phenomenological causality” does not demonstrate causality. It
describes a state of belief.

Warm wishes,

Ken

Professor Ken Friedman, PhD, DSc (hc), FDRS | University Distinguished
Professor | Dean, Faculty of Design | Swinburne University of Technology
| Melbourne, Australia | [log in to unmask] | Ph: +61 3 9214 6078 |
Faculty www.swinburne.edu.au/design

Terry Love wrote:

—snip—

In my previous email I was pointing to the possibility of a new area of
theory advance in design research that offers benefits over the existing
approaches.

I suggest there is a path to creating causal theories about designers
and users response to new and unknown designed outcomes based on
humans’ typical reactive responses in this situation.

This approach has the benefit of deriving testable causally-based
theories in their explanatory power go beyond the associatively-derived
theories about human responses to designs that currently dominate this
area of design theory. This latter follows from the general rule that
tested causally-based explanations are more useful than
associatively-derived information. This is because causally-derived
theories can predict outcomes for novel situations, whereas
associatively-derived information can only guide the design of things
that are similar or with incremental changes.

The underlying reasoning for the idea of creating causal explanations
about designers and users response to new and unknown designed outcomes
is straightforward: 

1. It is widely agreed that humans act on the basis of their previous
experience and learning combined with in-born responses.

2. This using of previous experience, learning and in-born responses is
evident in how humans respond to something new.

3. Designers create new and novel things.

4. Users have to respond to the new and novel things designers create.

5. Design researchers improve design processes and outcomes by
understanding how and why designers respond to the ‘new’ that
emerges in their minds and from their other design practices.

6. Design researchers and designers improve design processes and
outcomes by understanding the basis of how and why users respond to the
new and novel things designers create.

7. Causal explanations of the how and why designers respond to the new
provide the understanding of why these responses happen and thus offer
predictive power that can be used in other design situations.

8. Associative data about designers and users response to the new is
limited to providing information about how things have happened in
particular past situations.

9. Design theory development comprises formalising the above
knowledge.

10. Design theory development benefits more from causal explanations
than associative data.

11. Currently, the literature on these issues is dominated by
associative data (e.g. the excellent work by Vesna et al in QUT on
intuitive interfaces, Nielsen’s work on usability, the work on Design
and Emotion, the CSCW literature, the HCI literature...)

12. Individual human responses to the unknown are typified by fixed,
‘chunks of response triggered by their observation of aspects of
the new situation.

13. Evidence of the above is from multiple sources including research
findings relating to ‘knowledge chunking’, ‘fixed action
patterns’, ‘archetype use’, ‘value judgement’,
‘paradigms’, ‘disciplines and professional knowledge’ along
with everyday empirical experience.

14. The most obvious way, to me, to start to develop such a
causally-based subfield of design theories is to first map examples of
where users exhibit archetypical responses and then catalogue these
archetypes and responses in terms of set and mereological taxonomies.

15. These provide a conceptual basis for analysing how and why
designers and users respond using these archetypical responses, i.e. the
identification of a taxonomy of typical trigger factors.

16. The relationships between elements in taxonomies are the basis of
mid-level causal theories of how designers and users respond to the
new.

17. Analysis of the underlying foundations of the responses, archetypes
in these taxonomies provides an epistemological and ontological
foundation for such mid-level causal theories, and themselves offer
low-level causal theories of how designers and users respond to the new.
Almost certainly, these latter will be at the level of biological basis
of aspects of cognition, perception and emotion (biological basis of
embodied behaviour).

18. The development of effective design guidelines is restricted by the
types of theories about how designers and users respond to the new. The
use of theories based on associative data is limited to identifying
design guidelines of things that are similar, i.e. are limited in their
validity to incremental design (by definition).

19. The outcome is a new causally-based body of theory about how
designers and users respond to the new that offers a foundation for
better design guidelines that can address truly novel design rather than
incremental design. 

My apologies, I was trusting you would infer all the above (and more)
from my previous post on the help desk employees rant about user’s
‘superstitious’ thinking.

—snip—