of stock available in the city, balancing out such 
factors as the type of accomodation available, its size 
and location and cost against their own needs and 
resources. Finally there are the "controllers", the 
actors who make planning policy and manipulate the 
various sets of controls within which the providers and 
occupiers must operate, This account approximates more 
closely to the British than the American situation, where 
there would be other sets of actors representing 
politicians and pressure groups. 
Even at a very simple level, a view of the city which 
starts from the stock of structures soon unfolds into a 
most complicated pattern, which, like every other aspect 
of the city, is an extraordinarily rich mine of research 
themes and materials, A first task for any research 
worker is to narrow the focus and to identify problems 
which will be interesting and rewarding. 
The Limits of Our Study 
First of all let us limit our interest to the modern 
industrial city. The problem springs immediately into 
clearer focus. We see that the models of dealing with 
change in modern cities are very different from those 
adopted in other settlements and at other times. For 
example, the accomodation in an African village or 
encampment may be changed many times as population 
expands or as new social or family groups come into 
being. But the main pattern of accomodation 
remains unchanged. The temporary nature of buildings in 
African villages does not alter the permanence of the 
pattern. Buildings in these primitive societies are very 
simple and the same form, with minor variations is used 
for most activifies. In modern cities the permanence of 
buildings, combined with an increasingly rapid rate of 
change in activities, cannot be dealt with simply by 
rebuilding in the same manner as before. 
A focus upon the modern industrial city brings immediate 
advantages. But some difficulties arise from the 
complexity of the city. Two approaches are possible; to 
cover the total pattern of change in all its aspects; or to 
select a single aspect or indicator of change and to study 
it in some depth. 
At first sight the first alternative seems most attractive. 
Since the pattern of change is very complicated, and it 
is difficult to separate one aspect from the next it seems 
sensible to study a total pattern, even if this means 
restricting our study to a single part of the city. 
But there are difficulties. It is difficult to define or 
isolate an area of the city which has sufficient 
coherence to support exact hypotheses and which can be 
usefully related to other parts of the city. How should we 
know what to measure? And if adequate measurements 
were possible, how could we judge whether the answer 
were right or wrong? The more we think about the 
proposal to consider all aspects of change in a particular 
part or parts of the city the more we see that it is not 
susceptible to rigorous analysis, 
By isolatinga single facet for study across the 
whole face of the city we gain a number of advantages. 
First, and most important our studies gain in precision 
and clarity. Because a single topic is brought into focus 
it is possible to handle the problems with more rigour. 
Second, data problems become more manageable. Third. 
it is possible to formulate testable hypotheses 
about the behaviour of a single aspect of the city, and 
to decide with some degree of accuracy upon the 
correctness of the answers we obtain. 
The selection of a topic or facet for study becomes then 
of immense importance. What kind of criteria should be 
used to isolate such a topic? First, the facet chosen 
must be easily distinguishable from other elements in the 
patterns of urban growth. Second, it must be an 
important element in forming the general pattern of 
growth and change in the city. Third, there should be a 
well-documented history of policies and controls 
concerning the topic chosen. Fourth, there must be 
coherent and well organized data available. Finally, 
the facet chosen must be some kind of "major index" of 
urban growth. 
A number of topics fulfil these criteria, but one in 
particular stands out in the modern city - the growth of 
offices. Office growth is a major index of urban growth 
today. One of the main reasons for the continued 
existence of modern cities is because of the focus they 
provide for communications. There are many vehicles 
for such communications, but much of it is carried in 
pieces of paper which emanate from offices - the paper 
metropolis is indeed a major phenomenon today . 
The pattern of office growth in London thus fulfils all the 
criteria for our research programme. It has been importan! 
to London for many years, and the post-war era has been 
especially critical. Major policy decisions have been 
taken concerning office growth in London. A great deal 
of data is available from various sources concerning the 
pattern of office development in London. Finally, office 
growth can be distinguished from other facets of urban 
growth in London, and lend itself to treatment in model 
form. 
Having decided upon a topic for study, and a venue for 
investigation we should set some limits on the span of 
time we consider. The period for study is largely 
dictated by data resources. It is only since the Town and 
Country Planning Act in 1947 that office growth and 
development has been subject to control, and these 
controls have given rise to a considerable amount of 
useful information. There are also theoretical reasons for 
such a choice, The post-war years have seen a general 
increase in communications activities of all kinds, so 
that this period must be significant to the study of office 
accomodation and activity in London during the years 
since 1947 
A First Hypothesis 
Before a research worker can ask the right questions 
about his subject he must set up some hypotheses - 
however crude - about the processes he is looking at. 
Such hypotheses may be modified during the progress of 
the work, but they are an essential starting point. 
Consider the development of office activities as a birth- 
growth -death process. Each activity seeks different 
accomodation at different stages of its life cycle; 
balancing space, rent and location, against age and 
condition of structural stock, tenancy conditions, and so 
on. Activities are generated or "born" at a rather small 
size, and the circumstances of birth often affect their 
location. Most activities begin life in adapted premises, 
for it is here that they find cheap rents which they can 
ARCH + 1(1968) H.4