1. THE PROBLEM WITH
BUILDING USABILITY
(one of many)
+ a super-short crash course in Space Syntax
Jakub Krukar
Northumbria University, Newcastle, UK
jakub.krukar@gmail.com
2. Imagine a room.
And another room, next to it. One more…
Actually, just build a house in your mind.
Done? Cool, now mentally zoom out.
Look at it from the top, look at its floor plan.
You quite probably ended up with something similar to this...
4. I know, I know - not exactly the same shape, not exactly the
same size. But I bet the main idea is similar - a hallway, a centrally
located living room, a kitchen connected only to the living room
or to the hallway, etc.
The theory of Space Syntax looks at layouts emphasising
relations between spaces, not its 'visible' properties, such as size
or colour. If you ask a rich guy and a poor guy to design a house
for themselves, the size will be different, the number of
bedrooms and bathrooms will vary.
BUT they'll both have a very similar idea of what a house is:
where each of its parts should be located relative to the other
parts. None of them would put a toilet in the centre of the
house with 4 entrances connecting to it. Have you?
5. So let me just visualise these relations.
Again, that's our layout:
6. Let's divide it into sub-spaces: sub-spaces in which every person
is able to see every other person occupying the same sub-
space.The reason for this is that the concept of a 'room' can be
unclear sometimes and we need a more 'objective' concept.
We'll call it a convex space:
7. And now we need to turn this into a graph. So each space is a
node and every connection between any two spaces that can
be walked by a human is a link:
9. And justify the graph:
Got it!
That's the representation of the connections between spaces.
The reason why it's 'better' compared to a regular plan is that
it's independent of all those things that vary from house to
house (like the size of your bedroom) but allows us to catch any
repeatable patterns occurring in many different buildings
(such as the living room in the centre of it all).
10. We all have a more or less similar idea of what a 'house', a
'hospital' or a 'school' should look like. So you and me, we both
EXPECT to have a number of separate bedrooms in a house,
not all connected to each other (that would be a funny house, right?).
In Human-Computer Interaction, these ideas about how things
should look and work like are called 'mental models'.
But in an art gallery, for example, rooms often can be connected
to each other, allowing you to pass through intermediate rooms
without going back to a corridor.And no one seems to be
surprised about it!
That's because the mental model of spatial relations you're
using is context-dependant and will vary between places.
11. source:
this and the following
images are from
B.Hillier & J.Hanson -
The Social Logic of Space
12. And now the cool bit… (yeah, I think it really is cool!)
Space syntax is also used in archeology.
Why?
Because as it happens, these mental models can be compared
not only between contexts, but also between cultures.
13.
14.
15. And if we use our mental models to construct spaces
(like an architect designing houses - he uses the 'common idea of a house')
it means our architecture reflects those mental models.
"It is this ordering of space that is the purpose of building, not the
physical object itself.The physical object is the means to the end. In
this sense, buildings are not what they seem. [...] Buildings are not
just objects, but transformations of space through objects."
B.Hillier & J.Hanson -The Social Logic of Space
16. When looking at cities which
emerged 'naturally' - from being a
small village, through a town - it may
seem some of them look very
chaotic.
Both on a map and from the
viewpoint of a street.
BUT when we measure the relations
between spaces they happen to be
very regular, repeatable and equal for
everyone
(e.g. every house might have an access to the
main street which is not a cul-de-sac, a backyard
garden, and be surrounded by 3 neighbours).
17.
18. Sizes and shapes of these places differ (that's what makes it look
chaotic) but the main rules stay the same. So there's a hidden
order in this natural chaos - and we could assume that's our
'original' mental model here.
19. Look at gecekondus - illegal settlements inTurkey which evolved
with no urban planning whatsoever...
source: wikipedia
20. Try entering one of their small, curvy streets and you instantly
get lost. But to the people living there, who built these houses
and these streets it's all natural and somehow they never get
lost.
ATurkish friend even told me he never knows how to get to a
place but he always finds it.Ask him to explain the route though
and he'll have no idea.
source: http://inuraistanbul2009.wordpress.com/
21. If this is the 'natural order'
we'd like to replicate and
enhance, it looks quite
complicated, doesn't it?
source: wikipedia
22. So now, how can this poor, humble (khe, khe:) architect or
urban planner figure out what people want or what would fit
their needs? Well - that's a common issue in user-centered
design - you're trying to discover their implicit mental models,
behaviour patterns and adjust or fit your work into that...
In the architectural/urban context, if it's not done right, that's
when we get lost in a place - because we've been EXPECTING
something else in here. Maybe that's when a home doesn't feel
'homely' or when a neighbourhood turns into slums, too?
23. On the one hand, from this perspective, primitive settlements
reflecting natural patterns of use seem to work pretty well... In
which case maybe a top-down design is doomed even before
the first line in AutoCad is drawn?
On the other hand, all those researchers, architects and
designers do what they do for a living only because they believe
that something can actually be fixed.That planned intervention
can be (yet not always is) better than leaving things to take their
own course. But then - what do we really want to achieve
through these adjustments? And how do we decide which of
the possible solutions are better than the others?
24. JAKUB.KRUKAR@GMAIL.COM
This presentation was initially written for the guys from
Humans In Design. Check out their related blog post.
Thanks toTristan Cook and Prof. Ruth Conroy Dalton for their comments.