Hello,
What follows is a relatively long post about Design History that brings
together threads from a number of areas of design research.
If this is not your interest, you might find it saves time to stop reading
here.
Responses to my earlier post indicated I was wrong assuming design
research knowledge is shared across design fields.
I’d also assumed members of this phd-design list held a view of design as
‘domain independent’. That seemed to be a reasonable assumption for a
couple of reasons.
Firstly, the idea that design activity is intrinsically ‘domain independent’
is the original perspective of design research from the 1960s.
Secondly, many posts of members of this phd-design list strongly indicate
that they assume domain independency of design is the many posts that define
design generally (‘Design is ….’) rather than seeing design in a domain
specific say , e.g. ‘The word ‘Design’ in communication design means … which
differs from how it is used in XXX in the following ways…’.
Taking a ‘domain independent’ view of design necessarily implies
understanding of design theories across all design disciplines. On this
basis, I’d assumed that readers were aware of design research developments
and findings across all design fields.
The feedback indicated I was mistaken.
The feedback indicates it is more common on this list for researchers to be
domain-specific with their design research knowledge. It means that design
research knowledge is limited to what is available in the design domain in
which they work. There are many fields of designers, and different knowledge
about design from design research is available in each.
Then it becomes a problem of understanding when someone draws on design
research from different design disciplines , which is what I’d done.
A strong suggestion from one member of this list was that I post a more
fulsome explanation of the reasoning and assumptions on which my previous
post was based. This is that post.
Before starting , I’d like to comment that I support the idea that Design
History has some useful roles in design education.
HOWEVER, I’m suggesting that critical epistemological and operational review
of Design History does not support many of the existing claims for it. This
post details the reasoning.
More, such a review offers a clearer view of the functional limitations of
Design History vis-à-vis design practices and design education. It also
offers clarity over a broad raft of circumstances in which the claims for
Design History are unjustifiable and the findings and analyses of Design
History are faulty/mistaken.
Many of these problems with, and limitations of, Design History as it
currently is practiced apply importantly and particularly to design
practices and design education.
Perhaps the first assumptions I didn’t state explicitly in my previous email
and perhaps should have is,
‘The PRIMARY purpose for designers of ALL design theory and practice (design
research; design history, design science, design principles, methods and
practices; and all other theoretical aspects of design) is to IMPROVE THE
PREDICTION OF DESIGN OUTCOMES FROM DESIGN DECISIONS’.
Some argue that the above is the ONLY purpose of the theoretical side of
design: if you exclude reasons like getting fat salaries for design
researchers and design historians and the like (which is not a bad reason
but nothing to do with design per se).
From the above perspective, a key metric for Design History is how well
Design History acts to support prediction of design outcomes. (It’s
important to distinguish between design outputs and design outcomes.)
That is the first point.
Next.
The second thing I didn’t say and perhaps should have is that,
‘PREDICTION OF OUTCOMES RESULTING FROM DESIGN DECISIONS is central to, and
essential for, professional design activity’.
If one is working for a client and cannot predict the outcomes of one’s
design decisions and designs, then one is guessing: at the expense of the
client and of all the public and the environment that are affected by one’s
designs. This is behavior to which the old term ‘chancer’ applies (
<https://dictionary.cambridge.org/dictionary/english/chancer>
https://dictionary.cambridge.org/dictionary/english/chancer).
The primary reason we undertake design research and create design principles
and methods is to IMPROVE PREDICTION OF THE OUTCOMES RESULTING FROM DESIGNS
to reduce the unethical amount of guessing and chancing at the expense of
others to our benefit.
So, to recap to here, I’ve so far drawn attention to:
1. The primary role of ALL theoretical aspects of design (including
Design History and Design Research) is to provide better prediction of
design outcomes resulting from design decisions.
2. That PREDICTIONS OF DESIGN OUTCOMES is perhaps the most important
aspect of creating designs in a professional context.
Next.
The third thing I didn’t explain and perhaps should have is about feedback
loops.
‘A feedback loop occurs when different aspects of a design and/or its
context (e.g. users) influence each other such that the behaviour of a
causal factor is influenced by outcomes it causes - usually through other
factors.’
To say it a different way, Factor A in a design causes things to happen
including changing the behaviour of Factor B. However, the behaviour of
Factor B, perhaps via other Factors C and D, influences the behaviour of
Factor A, which then changes how it influences Factors , B, C and D and
hence influences itself, which continues the changes. This is an example of
a SINGLE FEEDBACK LOOP.
There are many examples of this single feedback loop situation in every
design field. The most common example that is used is the idea of a
thermostat. A heater (Factor A) influences the behaviour of the temperature
(Factor B) in the house and the behaviour of the thermostat (Factor C) which
in turn, when the temperature is too high influences the behaviour of the
heater (Factor A) by turning it down or off.
Typically, the limiting outcome of such a feedback loop is to result in the
design outcomes becoming either unstable or highly stabilized.
Next.
Some design situations can be very COMPLICATED but have NO FEEDBACK LOOPS.
In contrast, COMPLEX design situations have MANY FEEDBACK LOOPS.
This results in another assumption, conventional in systems theory, that,
‘IF A DESIGN SITUATION CONTAINS FEEDBACK LOOPS THEN PREDICTING THE DESIGN
OUTCOMES REQUIRES PREDICTING THE BEHAVIOURS OF THE FEEDBACK LOOPS AND CANNOT
BE DONE OTHERWISE’.
So, to recap to here, I’ve drawn attention to:
1. The primary role of ALL theoretical aspects of design (including
Design History and Design Research) to provide better PREDICTION OF DESIGN
OUTCOMES resulting from deign decisions.
2. That PREDICTIONS OF DESIGN OUTCOMES is perhaps the most important
aspect of creating designs.
3. FEEDBACK LOOPS are an important part of PREDICTING DESIGN OUTCOMES.
If a design situation contains feedback loops then PREDICTING THE DESIGN
OUTCOMES REQUIRES PREDICTING THE BEHAVIOURS OF THE FEEDBACK LOOPS.
Next.
The fourth thing I didn’t say and perhaps should have is that,
‘In design fields that specialize in addressing complex design issues, there
is a very careful distinction between SIMPLE, COMPLICATED, COMPLEX and
CHAOTIC design situations’.
The distinction hinges around the NUMBER OF FEEDBACK LOOPS and prediction of
the consequences from their behaviour.
SIMPLE DESIGN: low number of elements and linear relationships. Prediction
of outcomes is straightforward and can be done logically, using critical
thinking – in theory at least they can be done mentally ‘in mind’ or in some
cases by feelings or intuition.
COMPLICATED DESIGN: high number of elements and relationships with a maximum
of one feedback loop. Prediction of outcomes is again straightforward and
can be done logically, using critical thinking – in theory at least they can
be done mentally ‘in mind’ or in some cases by feelings or intuition – often
prediction activities are undertaken across multiple stakeholders as
‘co-design/participatory design.
COMPLEX DESIGN: any number of elements and TWO OR MORE FEEDBACK LOOPS whose
behaviours are intrinsically predictable using mathematics or physical
modelling. PREDICTION OF OUTCOMES IS IMPOSSIBLE TO DO MENTALLY ‘IN MIND’ OR
BY FEELINGS OR DISCUSSIONS BETWEEN INTERESTED PARTIES - but can be achieved
by modelling (usually mathematically).
CHAOTIC DESIGN: any number of elements and two or more feedback loops whose
behaviours are intrinsically unpredictable.
Next.
The fifth thing I didn’t say and perhaps should have is that in some areas
of design research (especially system design fields), there is awareness
that,
‘Humans are UNABLE TO PREDICT IN THEIR MINDS THE OUTCOMES OF DESIGNS WHOSE
BEHAVIOUR ARE SHAPED BY MULTIPLE FEEDBACK LOOPS’.
This is a biological limitation of human thinking/feeling/intuition.
The above is effortless to prove. The research method is to set up examples
of multiple feedback loop situations that we can exactly predict their
behaviour and future outcomes through modelling or mathematical
representation. Next is to ask humans to predict the outcomes of those
systems and ask how confident are they that they are understanding the
system and can predict the behaviour of the outcomes. Then, ask them to do
that prediction and then we compare their answers with the actual
behaviours and outcomes of the systems.
One of the classic documents identifying this limitation is that of Jay
Forrester in 1971 (well actually he seemed to have identified the problem in
the 1950s but the most accessible paper is 1971). (see Forrester, 1971).
The idea of ‘wicked problems’ was a later simplified derivation from
Forrester’s work applied to urban planning and, then, from the time in the
1980s when architecture and urban planning were the central discipline of
design research and Design Studies became adopted into design research.
Then, later, in the 2000s became adopted into Art and Design as these
design fields started to become involved in design research.
Forrester’s work is in many ways better reading and more useful than Rittel
and Weber or others in the Design Studies fields since Forrester (and later
Sterman) cover the territory in a much more advanced and sophisticated
manner compared to the thinking about ‘wicked problems’.
My own contributions to design research in this area (which are relevant to
this post about Design History) are:
1. Identifying that the human biological limitation is at the 2
feedback loop boundary.
2. Applying Forrester’s findings to design research and design fields
outside systems design
3. Identifying the implication from the above, that neither individual
designers, nor groups of stakeholders are capable of predicting the outcomes
of design decisions, in mind or in discussion, where design situations have
multiple feedback loops;
4. Identifying that, by implication, this challenges the validity of
co-design, participatory design, community participation in planning and the
like for complex and chaotic design situation. That is, participatory and
collaborative design or any process using one or more participants is
useless to try to predict design outcomes in multiple feedback loop
situations;
5. Identifying that designers and others self-deceive/delude
themselves whilst addressing multiple feedback loop design situations.
Designers (and design researchers) are unaware that they are functionally
unable using thinking or feelings/intuition to predict the outcomes of
design situations with multiple feedback loops and instead (and part of a
much bigger design theory problem) designers INCORRECTLY believe on the
basis of their internal perceptions that they understand the workings of
multiple feedback loops situations and are able predict outcomes.
6. Identifying that when designers and design researcher try to
predict the output of multiple feedback loop systems they fail and get
incorrect answers because they undertake one or more of the following
processes that result in compromised/faulty findings: a) they treat the
situation by ignoring any feedback loops; b) they convert multiple feedback
loops to a single feedback loop; c) they hide the feedback loops by blurring
the picture – usually by attempting stakeholder analysis or co-design or
some other consultative process; d) they try to identify a single outcome
(note: when there are multiple feedback loops the outcome is always dynamic,
changing over time); and/or, e) they attempt to force the explanation of the
behaviour of multiple feedback situations as a narrative or present it
somehow in written terms.
7. Identifying that in the design and design research literature the
only way to predict outcomes of design situations with multiple feedback
loops is via mathematical or physical modelling. Verbal and written
descriptions are fundamentally insufficient to explain outcome behaviours
except at a very superficial level.
Some of the above is described in Love, 2010a and Love, 2010b (see
references below).
So, to recap up to here, I’ve drawn attention to:
1. The primary role of ALL theoretical aspects of design (including
Design History and Design Research) to provide better PREDICTION OF DESIGN
OUTCOMES resulting from deign decisions.
2. That PREDICTIONS OF DESIGN OUTCOMES is perhaps the most important
aspect of creating designs.
3. FEEDBACK LOOPS are an important part of PREDICTING DESIGN OUTCOMES.
If a design situation contains feedback loops then PREDICTING THE DESIGN
OUTCOMES REQUIRES PREDICTING THE BEHAVIOURS OF THE FEEDBACK LOOPS.
4. Areas of design that focus on complex design situations have found
it useful to differentiate between SIMPLE, COMPLICATED, COMPLEX AND CHAOTIC
design contexts on the basis of feedback loops
5. That humans not only cannot think through or predict the outcomes
of complex situations involving multiple feedback loops. In addition, people
delude themselves that they can do so. Analysis of multiple feedback loops
systems results in faulty predictions when individuals or groups try to
insist that they can understand them and force the complex picture into less
complex pictures that can be addressed by the limited abilities and tools
of human thinking. That it is only possible to predict the outcomes of
multiple feedback loop situations via computer or physical modelling.
Next,
You might at this stage be asking what has this to do with Design Education,
Design Studies and Design History?
The sixth, seventh and eighth things I didn’t say and perhaps should have
are that,
‘Most of the SITUATIONS OF SIGNIFICANT INTEREST to predicting the outcomes
from DESIGN are MULTIPLE FEEDBACK LOOP SITUATIONS’.
This means that,
‘Most of the issues WORTHY OF INTEREST IN DESIGN HISTORY are MULTIPLE
FEEDBACK LOOP SITUATIONS’.
And, by implication from the above,
“The CURRENT METHODS OF DESIGN HISTORY, and its attempts to describe and
explain them in terms of language will always, FOR MULTIPLE FEEDBACK LOOP
DESIGN SITUATIONS, RESULT IN FAULTY, INCORRECT OR INACCURATE DESCRIPTIONS OF
OUTCOMES.”
The latter is easy to prove deictically.
However, if someone on the list has a proof that things are otherwise: that
current methods of Design History have the ability to predict the outcomes
of design situations with multiple feedback loops; I would love to read it.
And I’m happy to change my views.
So to recap the assumptions on which the contents of my previous email about
Design History was based,
1. The primary role of ALL theoretical aspects of design (including
Design History and Design Research) to provide better PREDICTION OF DESIGN
OUTCOMES resulting from deign decisions.
2. That PREDICTIONS OF DESIGN OUTCOMES is perhaps the most important
aspect of creating designs.
3. FEEDBACK LOOPS are an important part of PREDICTING DESIGN OUTCOMES.
If a design situation contains feedback loops then PREDICTING THE DESIGN
OUTCOMES REQUIRES PREDICTING THE BEHAVIOURS OF THE FEEDBACK LOOPS.
4. Areas of design that focus on complex design situations have found
it useful to differentiate between SIMPLE, COMPLICATED, COMPLEX AND CHAOTIC
design contexts on the basis of feedback loops
5. That humans not only CANNOT THINK THROUGH OR PREDICT THE OUTCOMES
OF COMPLEX SITUATIONS involving MULTIPLE FEEDBACK LOOPS. In addition, people
DELUDE THEMSELVES THAT THEY CAN DO SO. Analysis of multiple feedback loops
systems results in faulty predictions when individuals or groups try to
insist that they can understand them and force the complex picture into less
complex pictures that can be addressed by the limited abilities and tools
of human thinking. That it is only possible to predict the outcomes of
multiple feedback loop situations via computer or physical modelling.
6. Most of the situations of significant interest to PREDICTING THE
OUTCOMES FROM DESIGN ARE MULTIPLE FEEDBACK LOOP SITUATIONS.
7. Most of the ISSUES WORTHY OF INTEREST IN DESIGN HISTORY ARE
MULTIPLE FEEDBACK LOOP SITUATIONS.
8. The CURRENT METHODS OF DESIGN HISTORY, and its attempts to
describe and explain them in terms of language WILL ALWAYS, FOR MULTIPLE
FEEDBACK LOOP DESIGN SITUATIONS, RESULT IN FAULTY, INCORRECT OR INACCURATE
DESCRIPTIONS OF OUTCOMES.
In short, this indicates that conventional critical analysis and critical
historical analysis of Design History or otherwise are insufficient for
purpose in addressing complex design situations, i.e. those involving
multiple feedback loops.
This is regardless of whether such skills of critical analysis were actually
taught to a sufficiently high level in Design Education.
In conclusion, the above indicates:
Design History CAN work as a tool for helping designers predict the
outcomes of design decisions but only for SITUATIONS WITH LESS THAN 2
FEEDBACK LOOPS.
This can be done by simply copying/plagiarism.
It can also be done by building on knowledge and analyses of the past. This
latter is however subject to the limitations of design education in teaching
designers critical analysis to a sufficient professional standard, which
does not seem to occur (again I’m happy to be corrected with evidence).
The above also explicitly demonstrates that Design History DOES NOT work as
a tool for enabling designers to predict the outcomes of design decisions
for COMPLEX DESIGN SITUATIONS with 2 or more FEEDBACK LOOPS (i.e. most
design situations of interest).
Describing the analyses in words or even in drawings does not provide the
epistemological ability to enable such prediction.
One obvious limitation or a written explanation is the outcomes of
situations with multiple feedback loops are changing over time and the
changes depend on a range of factors that also change over time and any
representation in language that gives a fixed description is unable to do
this except in the broadest terms – which is hardly prediction.
Taken together these present a problem for designers and design educators
that hasn’t yet been addressed:
“How do you identify which design situations have multiple feedback loops?”
The most practical way to identify whether a design situation has multiple
feedback loops is to map out the causal relationships between important
factors in a design that shape outcomes. These immediately show the number
of feedback loops. The method is called ‘CAUSAL LOOP DIAGRAMS’.
These can be undertaken on paper or, more easily, one using a software tool
for creating Causal Loop Diagrams. Vensim is a well-regarded open source
tool that works on MacOS and Windows ( <https://vensim.com>
https://vensim.com). Such causal loop diagrams can be created by a single
individual or in a participator, collaborative multi-stakeholder
environment.
Identifying the kinds of feedback loops and the patterns of their
relationships can offer increased information through the use of ‘system
archetypes’. These system archetypes do not themselves enable prediction of
outcomes (that needs mathematical or physical modelling). However, the
system archetypes can give some idea of the types of behaviour that a
design situation is most likely to exhibit (
<https://thesystemsthinker.com/wp-content/uploads/2016/03/Systems-Archetypes
-I-TRSA01_pk.pdf>
https://thesystemsthinker.com/wp-content/uploads/2016/03/Systems-Archetypes-
I-TRSA01_pk.pdf).
These latter (systems archetypes) would be a really invaluable part of
Design History analyses if used.
In fact, they would help simplify and validate almost any Design History
analysis and provide a unifying and simplifying tool for explaining
findings.
Even better from the academic point of view of avoiding nonsense being
published, the combination of Causal Loop Diagrams and System Archetypes
provide an effective tool for reviewing the validity of Design History
publications.
From my own experience, the use of causal loop diagrams to understand design
situations appears non-existent in the Design History literature (I’d
welcome examples of its use in Design History).
What I hope to have outlined above is part of the background to my previous
email.
My apologies for my earlier email, when I presumed all this was obvious and
all I needed to do was to remind design researchers of the situation.
Most of the above, however, is already available can be found from
searching the phd-design list.
Best wishes,
Terry
==
Dr Terence Love,
School of Design and Built Environment, Curtin University, Western Australia
CEO, Design Out Crime and CPTED Centre
PO Box 226, Quinns Rocks, Western Australia 6030
<mailto:[log in to unmask]> [log in to unmask]
[log in to unmask]
+61 (0)4 3497 5848
ORCID 0000-0002-2436-7566
==
References:
J. W. Forrester (1971) Counterintuitive Behaviour of Social Systems.
Technology Review. Available:
https://ocw.mit.edu/courses/sloan-school-of-management/15-988-system-dynamic
s-self-study-fall-1998-spring-1999/readings/behavior.pdf
<https://love.com.au/docs/2010/Des-guide-gap-2feedbackloop.pdf> Love, T.
(2010). Design Guideline Gap and 2 Feedback Loop Limitation: Two issues in
Design and Emotion theory, research and practice. In J. Gregory, K. Sato &
P. Desmet (Eds.), Proceedings of the 7th Design and Emotion Conference 2010
Blatantly Blues. Chicago: Institute of Design and Design and Emotion
Society. Available:
https://love.com.au/docs/2010/Des-guide-gap-2feedbackloop.pdf
Powerpoint of presentation available https://love.com.au/docs/2010/TL-D
<https://love.com.au/docs/2010/TL-D&E2010.pptx> &E2010.pptx
Love, T. (2010). Can you feel it? Yes we can! Human Limitations in Design
Theory (invited plenary). Paper presented at the CEPHAD 2010 conference.
<https://love.com.au/docs/2010/CEPHAD-feeling-delusion.pdf> Available:
https://love.com.au/docs/2010/CEPHAD-feeling-delusion.pdf
PowerPoint of presentations available:
https://love.com.au/docs/2010/CEPHAD-TL-AA.pptx
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]>
Discussion of PhD studies and related research in Design
Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
|
