Dear Ward and Lars,
Thanks for your comments.
Yes, I along with many others in Lancaster in the 80s and elsewhere was
going down much the same route that pictures were helpful in thinking about
complex situations.
The problem is that we were naïve and confused merely 'complicated' with
'complex'. That was commonplace then. The world has moved on.
Yes, images help as an external memory store for thinking about
'complicated' situations. In effect, the 'story', 'narrative' or 'metaphor'
of the image acts in the role of a structured filing system. For these kind
of 'complicated' information situations (e.g. almost all of Bob Horn's type
of graphics) human memory and associated processes can extend quite easily
with a bit of external visual support.
'Complex' situations are different. Human cognitive and emotional biology is
not well suited to understanding or predicting the outcomes of 'complex'
situations.
Visual support offers almost no advantage. Humans appear to be unable to
predict the dynamic behaviour of systems with multiple feedback loops,
regardless of the visual support unless it is the output of a dynamic model
of the situation (in which case it is the model that is doing the predicting
and the human is simply observing).
All of the above are easy to demonstrate.
When discussing these issues, from experience it seems helpful to
differentiate between simple, complicated, complex, and chaotic situations:
1. Simple situations: low numbers of variables, degrees of freedom less
than or equal to the number of variables, no feedback loops
2. Complicated situations: High number of variables, degrees of freedom less
than or equal to the number of variables, none or one feedback loop
3. Complex situations: low or high number of variables, degrees of freedom
less than or equal to the number of variables, multiple feedback loops
4. Chaotic situations: low or high number of variables, degrees of freedom
greater than the number of variables, multiple feedback loops
Mostly, designers are pretty loose about their use of 'complex'. In the
discussions on this list, when claims are made for visual thinking applying
to complex situations the evidence given is usually visual representations
of 'complicated' situations. It’s a serious problem, an overclaiming, that,
uncritically the claim is extended to complex situations for which the
evidence does not apply.
Ward, the same distinctions and analysis apply also to linguistic issues.
Chinese pictograms are efficient for 'complicated' thought (i.e. lots of
factors, degrees of freedom of the situation less than the number of factors
(i.e. potentially predictable), and none or one feedback loop). Ditto Jung
and epistemic plurality.
Cheers,
Terry
-----Original Message-----
From: PhD-Design - This list is for discussion of PhD studies and related
research in Design [mailto:[log in to unmask]] On Behalf Of ward m
eagen
Sent: Friday, 16 April 2010 9:39 PM
To: [log in to unmask]
Subject: Re: Are visual approaches to design outdated?
Thank you Lars, this is a delicious insight, (at least for me, insights are
always relative) as I believe that what you are pointing to is that images
are somehow better at expressing relationships than words:
> “the complexity of human affairs is always a complexity of multiple
interacting relationships; and pictures are a better medium than linear
prose for expressing relationships.”
This makes sense, certainly when we are thinking about ‘new’ relationships
given the difficulty of language to express the unknown (thus its dynamic
nature, the use of metaphor, etc.) but then to connect that with the
difficulty in understanding the ‘true’ problem (as Bergson would say) from
the user’s perspective is an excellent subtlety.
> “There is little doubt that project requirements are the single biggest
cause of trouble on the software project front…”
We considered linguistic inadequacies in cognitive and analytical thought
from the perspective of creating in our article: Bauer, R. M. & Eagen, W. M.
2008. “Design Thinking — Epistemic Plurality in Management and
Organization.” Aesthesis, 2(3): 568-596, where we used the Jungian model of
psychological base functions for the conceptualization of human knowledge
rooted in epistemic plurality. Jung (1921) distinguishes four elementary
ways of knowing: thinking, feeling, sensing, and intuiting and we suggested
that design must necessarily consider all of these modes in creating. Your
insight that Epistemic Plurality, specifically visualization, must be
considered in the anchoring of the problem before the design solution can be
approached makes good sense, and I would offer this reference to you.
When Enerst Fenollosa wrote of the importance of the ideogrammic character
of Chinese ( Fenollosa, Ernest, Ezra Pound (1968) [1936]The Chinese Written
Character as a Medium for Poetry. San Franscisco: City Lights Books. ) he
was suggesting that the pictorial nature of Chinese writing lent itself to
complex thought.
“A true noun, an isolated thing, does not exist in nature. Things are only
the terminal points, or rather the meeting points of actions, cross-sections
cut through action, snap-shots. Neither can a pure verb, an abstract motion,
be possible in nature. The eye sees the noun and verb as one: things in
motion, motion in things and so the Chinese conception tends to represent
them.“ (Fenollosa, Pound 2001: 141)
The idea that linguistic based analysis has difficulty with the expression
of difference based in change gives credence to the suggestion that
visualization could provide more complex and creative results.
Lars, this is not offered as some sort of proof but only a path.
ciao,
we
On 2010-04-16, at 6:55 AM, Lars Albinsson wrote:
I just found a reference that are addressing this question:
“the complexity of human affairs is always a complexity of multiple
interacting relationships; and pictures are a better medium than linear
prose for expressing relationships. Pictures can be taken in as a whole and
help to encourage holistic rather than reductionist thinking.” ( Checkland,
1999 p A16)
In the field of IS there are a number of studies concluding that failures
are often correlated to problems with how different people put forward and
understand requirements in relation to a designed solution:
"There is little doubt that project requirements are the single biggest
cause of trouble on the software project front. Study after study has found
that where there is a failure, requirements problem are usually found at the
heart of the problem". (Glass, 1998 p 21).
I have not found any specific study that connects these two statements, but
doesn't it appear to be a possibly fruitful hypothesis? Perhaps HCI people
and others can provide studies that demonstrate that projects using
prototypes are more successful that others? That would be a fairly solid
argument for visualization being useful, at least in some ares.
Best Regards
Lars
Ps it would be interesting to learn if there indeed are non-visual
approaches in for instance architecture?
Checkland, Peter. (1999). Soft systems methodology : a 30-year retrospective
; and, Systems thinking, systems practice ([New ed.] ed.). Chichester: John
Wiley.
Glass, R. L. (1998). Software runaways: Lessons Learned from Massive
Software Project Failures Upper Saddle River, NJ: Prentice Hall PTR.
.........................................................................
LARS ALBINSSON
+46 (0) 70 592 70 45
[log in to unmask]
AFFILIATIONS:
MAESTRO MANAGEMENT AB
CALISTOGA SPRINGS RESEARCH INSTITUTE
UNIVERSITY OF BORÅS
LINKÖPING UNIVERSITY
.........................................................................
9 apr 2010 kl. 15.47 skrev Birger Sevaldson:
Hi
There's no evidence of any accurate prediction of dynamic design outcomes
visual or not. There's to my knowledge no such thing as accurate predictions
in design, or the world would be easy.
I am working with visualisation to help design for very complex situations
and to understand systems from a designers perspective and to use a
designerly approach to systems thinking.
In this process we map very complex information in big maps and we call them
GIGA-maps. It's a way of researching and learning a complex field and
patterns of relations in this field, more geared towards a holistic
understanding of the field than defining borders and hierarchies. We work
with information visualisation people. Typically they are geared towards
communication and representation and not towards using visualisation as a
process tool for individual designers and for co-design or for communication
of processes to stakeholders. The way the information is visualised and
presented is important for the process and the result.
Is there any proof that it works? No. Is there experience that it works?
yes.
Visualisation is crucial in any other field that deals with complexity. And
so is interpretation. A friend of mine did very advanced 4D voxel
visualisations of fluid dynamics and they discovered certain phenomena in
the visualisation. But these were patterns only to be recognised and
interpreted by the human eye and mind and visual intelligence. You need an
ostensive interpretation of the visual material in any case. Though the data
is quantitative the phenomena are only discovered when the data is
visualised and the proof is based on qualitative interpretation.
So Terry to my mind your question of proof is sort of on the hard end while
the systems thinking we do in design or even design thinking in practice in
general is on the soft end.
I think I said pretty much the same as Chuck formulated so elegantly.
Best
Birger
-----Original Message-----
From: PhD-Design - This list is for discussion of PhD studies and related
research in Design [mailto:[log in to unmask]] On Behalf Of Charles
Burnette
Sent: 9. april 2010 15:16
To: [log in to unmask]
Subject: Re: Are visual approaches to design outdated?
Terry, et al
I think the problem with complex systems science comes down to the
interpretation of results. There is of course the problem of nominal
interpretation when defining the input to any statistical method
(Garbage in-Garbage out). More importantly, it seems to me, is the
fact that you must adequately account for human interpretation of the
outcomes from systematically proscribed analysis, whether involving
multiple feedback loops or not. While, such feedbacks might or might
not have some corrective validity, it comes down to the question of
whether you can recognize the relevance of the outcome to the
variables you can act on and that you can believe in. This, puts it
back into the realm of communication between active stakeholders and
whatever shared understandings they can reach. It is ultimately a case
of human interpretation.
Similarly, the act of visualization or its almost instantaneous
interpretation requires the construction or deconstruction (I prefer
the words "blending" and "unblending") of recognizable elements and
relationships . The resultant mental image is a coherent (not just
visual) network of contributing information. It is holistic only in
that it has referential objectivity that affords opportunities for
selective interpretation and transformation.
In sum, it doesn't make sense to deny the human limitations in dealing
with information - whether posited through systems or other
recognizable forms. To much belief in nominal processing doesn't cut
it. Don't drink that kool-aid!
Chuck
On Apr 8, 2010, at 7:53 PM, Terence Love wrote:
> Dear Meredith,
> Thank you for your message. You express exactly the general state of
> mental
> play of the design field on this issue .
> " Visuals ... have a huge potential when we are talking about
> understanding
> complex problems" and "... *may* [my emphasis] be more than just
> representations", Visuals *may* "... have the potential for predicting
> directions or outcomes"
>
> This has been the speculation for decades. All of these are
> tentative and
> appear as yet to be without evidence that visuals can actually do
> these
> things. I'm wondering if they are simply wishful thinking and false
> claims?
>
> A test: "In complex design situations (multi-feedback loop and more
> than 3
> dimensions) is there ANY evidence visuals can accurately predict the
> behaviour of dynamic design outcomes?
>
> If anyone has any evidence, I'd be very interested to see it as it
> would
> open up a significant new area of design education and practice. It
> would
> also open the door to Visual Design being able to start to justify
> claims
> that design thinking approaches apply to more complex areas such as
> business
> strategy and other real world complex design situations.
>
> Best wishes,
> Terry
>
> -----Original Message-----
> From: PhD-Design - This list is for discussion of PhD studies and
> related
> research in Design [mailto:[log in to unmask]] On Behalf Of
> Constanza
> Miranda
> Sent: Friday, 9 April 2010 1:03 AM
> To: [log in to unmask]
> Subject: Re: Are visual approaches to design outdated?
>
> Hi Terence,
>
> Actually I'm doing research in this area. Visuals, which is one of the
> knowledge evidence in design research, in my opinion, have a huge
> potential
> when we are talking about understanding complex problems regarding the
> design interaction not just in the computer, but as Meredith Davis
> mentions,
> outside of them. Complexity regarding: access to particular
> communities, a
> diversity of disciplines working together or just networked
> individuals.
> This is the new scenario that design educators should acknowledge.
> Something
> more than a "trend", but a new conception of the discipline. We can
> go back
> to Sapient's experience models, that though, they were born for rapid
> business cycles, they can be a referent for the applied world in
> complex and
> interdisciplinary areas like public service. Visuals may be more
> than just
> representations. They might be used as models [simulations or
> theoretical],
> something that Hugh Dubberly has been addressing since around 2001,
> if not
> earlier. In this form, they have the potential for predicting
> directions or
> outcomes and, depending if the analysis is driven by statistical or
> more
> qualitative approaches, they might be more or less reliable.
>
> Best,
>
> Davis, Meredith
> 2008 Toto, I'Ve Got a Feeling we'Re Not in Kansas Anymore.
> Interactions
> 15(5):28-34
>
>
>
> On 8 April 2010 09:51, Terence Love <[log in to unmask]> wrote:
>
>> Hi David,
>> First a big thanks for the pointer to Bob Horn's work.
>> The fundamental issue to be addressed in any kind of design (and
>> often not
>> addressed by graphic designers) is accurately predicting the
>> behaviour of
>> the outcome. For example, how does one know if a poster and a public
> health
>> promotion program will achieve its aim of reducing smoking by 50%?
>> Why did
>> or didn't the design produce the right outcome? To ask these
>> questions
>> about
>> behaviour is central to design as a profession and to development of
>> suitable design methods, design processes and design education.
>> Many early
>> engineering design methods to predict behaviour were visual. With
> increased
>> complexity, they ran out of steam which is why engineering
>> designers now
>> use
>> different methods that primarily use visual for input and final
>> presentation.
>>
>> Bob Horn's work is tremendously useful and interesting. Thank you for
>> reminding me about it. I remember his hypertext book in the late
>> 80s.Bob
>> Horn's visuals and visual language approach however are not about
>> predicting behaviour of designs in the manner of a design method, nor
>> primarily about predicting the outcomes of complexity (many feedback
>> loops).
>> Their main role is knowledge mapping to make complicated
>> information
>> (more
>> bits) easier to access and think about. They do this in part by
>> acting as
>> an external memory store in the manner similar to that proposed by
>> Tony
>> Buzan some years ago with mind maps. They are primarily visual
>> representations of knowledge content rather than design methods that
>> predict
>> behaviour. Similar, though less attractive approaches were
>> developed in
> the
>> 60-s and 70s in the 'for beginners' comic books (e.g. Foucault for
>> Beginners
>> and Marx for Beginners etc), soft systems CATWOE maps, criteria
>> maps(which
>> echo critical path maps), logic maps, decision tree maps, group
>> decision
>> making support diagrams and flow diagrams.
>> Horn's description of the Cognome visual language project and
>> problems it
>> faces, identifies the same limitations and problems as above
>>
>
http://stanford.edu/~rhorn/a/recent/VUscnrioVisualLanguage.pdf<http://stanfo
> rd.edu/%7Erhorn/a/recent/VUscnrioVisualLanguage.pdf>except he
>> hasn't yet added the behaviour prediction problem as his aims as a
>> design
>> method don't reach that far... yet. It is, however, implicit in
>> his top
>> long range goal.
>> For improving the quality of design activity, the paper on his site
>> that
>> really made me sit up was the one of images that identify what we
>> don't
>> know
>>
> http://stanford.edu/~rhorn/a/recent/artclUnknowns.pdf<http://stanford.edu/
> %7
> Erhorn/a/recent/artclUnknowns.pdf>
>> A paper that might be useful for design research PhD students is on
>> how to
>> conduct research to get little or no effect
>>
>
http://stanford.edu/~rhorn/a/recent/artclHowtoGetLttleorNoEffc.pdf<http://st
> anford.edu/%7Erhorn/a/recent/artclHowtoGetLttleorNoEffc.pdf>
>> Amazing. Horn's recent papers and diagrams are at
>>
>
http://stanford.edu/~rhorn/a/recent/<http://stanford.edu/%7Erhorn/a/recent/
>>
>> Best wishes and thanks again,
>> Terry
>>
>>
>> -----Original Message-----
>> From: PhD-Design - This list is for discussion of PhD studies and
>> related
>> research in Design [mailto:[log in to unmask]] On Behalf Of
>> David
>> Sless
>> Sent: Thursday, 8 April 2010 3:14 PM
>> To: [log in to unmask]
>> Subject: Re: Are visual approaches to design outdated?
>>
>> Terry et al,
>> A provocative question. I would suggest a resounding NO. I think we
>> are
>> just
>> beginning to scratch the surface of possibilities in visual
>> approaches.
> The
>> work of Bob Horn just mentioned is an example of this. I think we
>> are at a
>> very early but highly productive stage in this kind of work. Look at
>> argumentation mapping as another example. Or the growing uses of MRI.
>>
>> BTW, in talking about complex systems, stakeholders, relationships,
>> feedback
>> loops and multidimensional problems you are using some of the prime
>> metaphors of our time.
>>
>> David
>> --
>>
>> blog: www.communication.org.au/dsblog
>> web: http://www.communication.org.au
>>
>
>
>
> --
> Constanza S. Miranda M.
> PhD Student NCSU-Design
> www.innovacionsocial.cl
>
> "Develop Design, Design to Develop"
|