Dear Peter - If you go back to the first paper on spatial integration
and movement (Hillier et al: 'Creating life: does architecture
determine anything' in Architecture and Behaviour 1987 - now defunct)
we showed that in certain kinds of complex social housing estate with
low intelligibility you found little correlation between the internal
structure and movement. What seemed to be happening was that in the
overly complex internal layout, people were moving to the outside by
the simplest route, and this created a pattern in which movement was
strongest at the edges and fell with line depth into the interior.
You can find an instance of the in an image in chapter 5 of Space is
the Machine.

Another relevent result was from Ruth Conroy Dalton in her doctoral
thesis, where she showed that subjects found it much more difficult
to navigate in an unintelligible immersive world than in an
intelligible one. All she did was take a set of blocks with
relatively strong linear relations between blocks and shift the
blocks slightly so that lines were impeded.The small changes made a
substantial difference to how well subjects were able to solve
navigational problems (see attached .ppt). Recently, we have shown a
parallel effect using forward sighted agents with Alasdair Turner's
DepthMap software. The traces of 10000 agents in an 'intelligible'
layout more or less followed the pattern of visual integration, but
with the blocks slightly moved the traces went all over the places
and tended to cluster in large convex spaces.

Overall, we expect to find good correlations between spatial
integration and movement in urban areas where the land use pattern
has been shaped over time by the natural movement pattern - for
example, shops have tended to go to location where the grid has
created natural movement. In implanted developments where this has
not happened, and where land uses have perhaps been assigned on
different criteria, we do not expect to find a strong natural
movement pattern, and other factors need to be modelled to understand
the movement patters. Space Syntax Limited have done a lot of work on
this with the Walkability Index which builds many other factors into
the regression model - though not into the spatial model as we find
it more useful to keep the effects of spatial configuration clear by
kepping them independent.

All this may not be relevent to your particular case. Do you have an
image of the layout, movement patterns, and land uses. It is hard to
say much that is useful with seeing what the settlement is like.

Anyway, look at the attached .ppt, taken from a recent lecture of mine. - Bill

>In writing this I'm hoping to get some feedback on some research that I
>am currently doing in the Canadian Arctic. During the 1950's, Inuit were
>moved from traditional camps into settled communities across northern
>Canada. The spatial layouts of these communities appear to have been
>determined primarily by servicing and administrative reqirements
>associated with the unique conditions of arctic environments. For
>example, because sewage and water have to be trucked to and from houses
>(permafrost prevents standard utility hook-ups) the street grid has to
>be made as compact as possible. This is done to minimise servicing
>costs. In addition, expansion of these settlements is often determined
>by the presence/absence of solid surfaces (gravel beach ridges versus
>softer tundra).
>
>I became interested in using axial analysis to better understand the
>configurational properties of these unusual settlements which appear to
>be unlike towns found in other areas of North America. Not suprisingly,
>many of the arctic communities I analyzed have produced low correlations
>between connectivity and global integration, indicating low
>intelligibility. A few summers ago, I began to collect observational
>data on pedestrian and vehicular movement in one Inuit community on the
>west coast of Hudson Bay. To my surprise, while vehicular movement (all
>terrain cycles and trucks) was highly correleated with local
>integration, this was not the case for pedestrian movement, where
>correlations were much, much lower. Given that all of the observational
>studies I have come across in space syntax research indicate that a
>strong correlation exists between pedestrian movement and integration in
>western cities such as Atlanta and London, I am interested in
>understanding why I did not observe this in the Canadian Arctic.
>
>My first thought is that it might have something to do with the fact
>that the occupants of these communities (Inuit) might be transposing a
>"uniquely Inuit" pattern of movement onto the communities in which they
>live....communities which reflect Euro-Canadian concepts of 'community'
>and adminitrative control.
>
>This research is still in its early stages and my observations are
>limited to a 3 month period in a single community. However, I was
>wondering if anyone had any ideas as to why these correlations appear
>lower than those observed elsewhere. Is human movement
>culturally-mediated? and if so, does this mean that arctic communities
>may not be spatially sustaining Inuit cultural values?
>
>cheers,
>Peter
>
>
>Attachment converted: Macintosh HD:pcdawson.vcf (TEXT/ttxt) (0008C2BB)