Pre-design of signage and orientation systems in Portuguese hospitals

E. Côrte-Real, A. Ribeiro, S. André, S. Batel, C. Lisboa, C. Rosa

UNIDCOM/IADE, Portugal
aiso@iade.pt

1. Introduction

Orientation, the product of the user behaviour and ambient, to be successful involves to
know where you are, to know your destination point, to know and follow the best path
for your goal destination, to be able to recognize your destination after you arrive to your
goal and then to able to came back to your starting point. (Carpman & Grant, 2002).
For all this questions we must considerer the architectonic and ambient configuration
to help user orientation and circulation. But this is not enough. The user needs to have
a space intelligibility revealing a clear relation between space functionality and cognitive
mapping.
To improve the quality of way finding in hospitals we have centred our research project
on a multidisciplinary approach of space, a sum of methodologies from other disciplines,
regarding improvement of the individual experience in space.
We gathered methodologies from Social Psychology, Cognitive Ergonomics, Classical
Physics and Space Syntax aiming to achieve a comfortable way finding in complex struc-
tures with a specific social meaning and without a program of signage - Portuguese Public
Hospitals. Although all this methods are independent fields or disciplines, gathered by
the design problem, became, as whole, a design research methodology. In fact, that deter-
mines that cognitive ergonomics is useless without Space Syntax when you think not of
orientation but in making an intervention on orientation. This means that the necessity
of producing design objects or a design system triggers the linkage between the disciplines
otherwise; design wise useless.
Design projects have several levels of complexity and they demand a series of phases.
From the conception up to the achievement, passing through study and experimental
phases to development, Design confirms the level and the capacity of answer of the project
to the identified need.

2. The Hospitals

Besides Hospital Santa Maria, Lisbon, a large health care facility, of German Design and
an University Hospital built in the 1950’s, we have choose two different buildings: Hospital
de S. José, Lisbon, a former Catholic Convent, also linked to University, adapted since
the 1830’s; Hospital da Cova Beira in Covilhã, 1998. Such structure is also a University
Hospital. These three structures represent, each, a specific kind of building used for these
facilities in Portugal. The symmetric building, Santa Maria, the cluster organization, S.
José and finally the new cellular structure used in the last decade (Cova da Beira)
The largest hospital in Portugal doesn’t have signage. Santa Maria’s Hospital, which
integrates also the Lisbon University’s Medicine Faculty, doesn’t have an active and co-
herent signage system. Most of the signs that remain visible are from the late 1950’s when
460 E. Côrte-Real, A. Ribeiro, S. André, S. Batel, C. Lisboa, C. Rosa

the hospital was built, or are ad-hoc photocopies.

Santa Maria’s is therefore not only an urgent case but also an extreme field of research
in way finding. It gives us the opportunity to study a system that works without a system,
in which comfort in way finding should be an important issue.

3. Theoretical Model

We propose a theoretical model based on the concept of Viscosity; using an interdisci-

plinary approach, defined through Design Theory, Social Psychology, Physics, Cognitive
Ergonomics and Space Syntax, that provides a link between architecture and signage
(space syntax); the access to way finding strategies (cognitive mapping); the signage im-
plementation (space syntax and cognitive ergonomics); and the definition of rules and
guidelines to improve signage (space syntax, social psychology, cognitive ergonomics and
design).
The analogy with this concept determines that low viscosity means that way finding
is uncomfortable and a lot of energy is spent in people’s movement. The use of physics
parameters to quantify the quality of fluxes in different built structures will allow us to
predict and experiment the efficiency of changes in the sign guidance system.
Using the Principle of Dimensional Homogeneity in physics, we adapted the formula of
viscosity, using the parameters measured concerning the quality flux of people to compute
a valid value of viscosity for each origin-goal section inside the building The proposed
methodology admitted two parallel approaches to define the different variables that mea-
sure viscosity values:

3.1. Users analysis

We made a query1 at destination points that inquired about: socio-characterization; dif-
ficulties in way finding; previous experience; relation between anxiety and interpretation.
The queries were made at destination points, in waiting rooms during the time for
external consultations. The sample is, therefore, limited to people that use this kind of
service.

3.2. Space analysis

For space analysis we have used the values of integration and connectivity of the buildings.
Space syntax has provided the basis of the spatial and structural analysis of hospitals,
which has permitted us to determine the paths characteristics and the user’s way finding
behaviour, as well as the decision points and the more or less segregated areas. (Fig. 185)

1 The Query: We have divided the query in seven groups. The first group analyses the previous experi-
ence of each agent before the present visit to the hospital. Such experience can consist, for example,
in previous visits to the hospital. The second group is an established and certified method to measure
anxiety. The third group intent to measure the difficulty and mistakes while walking a path ”A to
B”. That must be made at B after the path is completed. The next group studies the orientation
methods used by people. The group five studies the impact and use of the sign system. The last two
groups analyse the capacity to interpret visual and written information and willingness to read and
interpret the signs in a limited time interval. (figure 189)
 Pre-design of signage and orientation systems 461

Figure 185:

3.3. Classical Physics

This discipline provides us a formula (viscosity) that can be compared to the phenomena
of comfort in way finding.
Viscosity relates stress, velocity (space vs. time) and weight on a fluid. Stress resulting
from the calibre of the path is substituted by integration values resulting from Space Syn-
tax analysis. Distance is the result of connectivity (sections with no decision making) also
from Space Syntax analysis. Time is the difficulties declared by users. Weight is the abil-
ity/disability to interpret written or visual information. This formula is used to measure
present situations and quantify alterations determined by the placing of information.
Viscosity for liquids relates shear stress and the velocity profile of a fluid under such
stress. The absolute viscosity of a fluid is given by:

F/A = η.v0 /d (21)

(1) - Absolute viscosity
In the metric system absolute viscosity has units of kg/(m ∗ sec). In order to obtain
the equation for the viscosity in a building structure the principle of homogeneity can be
used.
First, the variables of which viscosity is a function must be identified: Suppose a certain
corridor from two places, A and B, in a general architectonical space. The flux of people,
agents, crossing that corridor characterizes its viscosity. That is, the easiest is the flux the
smaller is viscosity.
Notice that, as we shall see, the average velocity will be a function of the number of
times an agent stops to gather information about directions using several methods, in
order to reach a certain goal.
Therefore, the flux of people can, in first approach, be considered like a fluid, where
the “particles” are actually connected by information exchange necessity. Connections
between “particles of the fluid” are established, for example, when a person stops to ask
another for directions.
Friction with walls here not considered will be related with the necessity of observ-
ing the architectonical space. The least knowledge about the space demands much more
462 E. Côrte-Real, A. Ribeiro, S. André, S. Batel, C. Lisboa, C. Rosa

stopping time in observing signs on the surrounding space.

Due to queries characteristics we are unable to determined the exact path used by an
agent when walking from A to B. Therefore, the results observed in the query for a certain
origin-goal section from A to B are regardless of all possible paths to reach B from A.
We considered only the ideal path (the one of least distance) and all others are errors
decreasing the average agent velocity.
The viscosity parameter, η, therefore, characterizes the building structure and is not
a way to choose between two possible paths from A to B.

3.3.1. Relativistic Width of a corridor as a function of Space Syntax Concepts

If we think in terms of the width of the corridor we immediately realise that, as we decrease
d, we increase viscosity, since the components travelling there will have a higher chance
of colliding, that is, the flux decreases.
Yet, such is not a realistic view of an architectonical space. The widths of corridors
are not limits, except in emergency situations or jams, to the agents’ movement.
Nevertheless, it is possible to define a “relativistic width” (drel ) dependent of the real
spatial parameters that limit fluxes.
Such quantity shall be computed using spatial syntax analyses. We shall define that
the relativistic width is a direct function of the “integration” of the line, after the space
syntax analyses of a certain space is made. In short, more integrated lines are wider than
less integrated lines.
In order to have a normalised quantity for different buildings we shall divide the
line integration by the maximum integration value present in the architectonical space
(Relativistic width):

drel = drel (lineintegration/maximumlineintegrationvalue) (22)

- Relativistic width
We can establish that (Relativistic width):

drel = (lineintegration)/(maximumlineintegrationvalue) (23)

- Relativistic width
We can, therefore, see that one of the variables of which the viscosity depends upon is
the width of the corridor, d.

ηrel = ηrel (drel ) (24)

- Width d as a parameter
For a A to B section we use the average integration value of the lines present in the
ideal path divided for the maximum integration value in the system.

3.3.2. Agents “Mass” as a function of Agents Error probability. The relativistic charac-
teristic of viscosity.
We are thinking in terms of agents, in this case, people. If we think of people walking in a
building we would not, except if they were running, state that the most important factor
for its speed is its weight.
 Pre-design of signage and orientation systems 463

The importance of the actual racing speed would be an important factor studying
emergency building evacuations, which is not the case here. In our study, and using prac-
tical observations we have come to the conclusion that it’s the knowledge and psychological
characteristics that determines its capacity of information interpretation, which we asso-
ciate with its “mass”.
That is, mass will be measured by the stated error probability of a certain agent in
finding its correct root inside a building structure. Since we are researching with the goal
of improving the role of sign systems, the mass that will be computed results from the
test on interpreting visual and written information (Variables of the relativistic viscosity
function).

ηrel = ηrel (mrel , drel ) (25)

- Variables of the relativistic viscosity function
Where mrel is defined by (Relativistic mass as a function of error probability):

mrel = mrel (Errorprobability) (26)

- Relativistic mass as a function of error probability
If we think in terms of viscosity we realise that this is an important parameter for
computing viscosity since takes in account the properties of the “fluid” we are considering.
The error probability of the agents and, therefore, its “mass” increases with lack of
capability to interpret both visual and written information. Such factor shall be considered
when computing the agents’ mass.
In practical terms, the mass will, therefore, be computed using a query independent of
the surrounding architectonical space. The data to compute mass are obtained from the
last two groups of the query (groups and of the query).

mrel , = mrel (visualandwritteninterpretation) (27)

- Relativistic mass
In practical terms, in order to obtain a value for an agent ”mass” we shall use the
relation:

mrel , =< visualerrors > + < writtenerrors > (28)

To avoid null mass we add one:

mrel , =< visualerrors > + < writtenerrors > +1 (29)

To have the mass measured in quilograms we multiply all by

1kgmrel , = [< visualerrors > + < writtenerrors > +1].(1kg) (30)

3.3.3 The average Velocity as a function of Space properties.

The agents’ properties have been characterized by its relativistic mass. But space inside a
building has inner properties that must be taken in account. Namely, in a large building,
signalisation, architecture and other factors will determine the flux velocity. Therefore, one
parameter, capable of measuring these architecture characteristics is the ”average velocity”
of the agents. The data to compute time in velocity are obtained from the third group of
464 E. Côrte-Real, A. Ribeiro, S. André, S. Batel, C. Lisboa, C. Rosa

the query related to the average velocity measured in the field crossing the shortest path
with ;
The average relativistic speed of an agent will be computed by (Relativistic Velocity
of agents):

V rel = Vmax − n∗ (Vdecrease ) (31)

- Relativistic Velocity of agents
Both Vmax and Vdecrease can be experimentally determined by field queries.
Actually, in practical terms, people will consider a stop every time they stop, whatever
the reason for stopping, thus this is a good measure for their average speed. To com-
pute Vdecrease , we count all the stops due to using wrong routs, requiring others help for
orientation and observing surrounding architectonical space for orientation.
Notice that this direct relation with stoppage time and velocity could not be so simple
if the path size would tend to infinite. In such case only a formula relating total time and
stoppage time would be possible.
Also, the stoppage time is an average time. Situations like someone deciding to camp
in the middle of corridor could not be taken in account!
Acceleration is also not important to consider since the values for the velocities are
small enough quantities that makes acceleration almost instantaneous. For us, the agent,
either moves with a constant speed or is stopped.
For space in velocity we choose the total intersections at on path that result from
its connectivity. That means that we didn’t considered the real length of a path but the
number of sections with out intersections and, therefore, with no possibility of error. We
have, in conclusion seconds expressed by the average time plus the stoppages and meters
expressed by the connectivity.
We can, therefore, consider, velocity, as an important factor for computing relativistic
viscosity (The average speed as a parameter):

ηrel = ηrel (mrel , vrel , drel ) (32)

- The average speed as a parameter

3.3.4. A Formula for measuring the Architectonical Relativistic Viscosity using the Prin-
ciple of Homogeneity of Equations in Physics
From previous experiences, it is known that:

F/A = η.v0 /d (33)

From this expression we verify that, in the metric system, absolute viscosity has units
of kg / (m*sec) (Units of η):

[η] = Kg.m3 .s−1 (34)

Therefore, any expression to determine η must have these units as a result. We have
already identified the variables of which ηrel depends upon, so we can write (Variables of
the relativistic viscosity function):

ηrel = ηrel (mrel , vrel , drel ) (35)

 Pre-design of signage and orientation systems 465

- Variables of the relativistic viscosity function

According to the analyses made, it is now possible to write a general expression for
ηrel that, applying the Principle of Homogeneity of Equations in Physics will allow the
determination of the dependency of ηrel relatively to all the variables of which it depends
on.
Knowing the dimensions of each one of the variables it is possible to determine the
explicit dependency of ηrel relatively to each one of the variables (Relativistic viscosity
Equation simplified).

ηrel = vrel .mrel /d2rel (36)

- Relativistic viscosity Equation simplified
This formula allows the computation of viscosity of any corridor of constant width d,
for a person of relative mass mrel , travelling at an average speed v0 .
If more than one person is walking trough the corridor, the viscosity must take it in
account.
It can do so by computing the average for all terms for example (Average Relativistic
Viscosity Equation):

ηrel =< vrel > . < mrel > /d2rel (37)

- Average Relativistic Viscosity Equation
Where < mrel > is the average relativistic mass of all people and < vrel > is the
average speed of such persons.

3.3.5. Proposals for Path Choice Optimization

Distance is, as we have seen, the only parameter, in the described conditions, that actually
matters to optimize the effort of reaching a location inside a building structure. Therefore,
if one thinks of creating a signalization system to optimize traffic it must have, as goal,
taking in account all error factors previously described, diminish the average distance to
reach from any point to any other point in the building using the shortest path possible.
This conclusion does not imply that all intersections must have a sign indicating all
possible locations. Such would create a major confusion, at least. An optimal state of
signalization placing and spatial location must be determined. In a first analyze it seems,
from the observed above, that signs must increase viscosity and approximate the paths
chosen using the viscosity concept, to those chosen using the least action principle.
It is now possible to compare action and viscosity values to check if the least action
path is also the one of least viscosity.

4. Testing the Theoretical Model

When we observe the values, we can, immediately, notice that the lowest values of vis-
cosity are observed at Santa Maria. For a population with more difficulties in interpret
information (mrel ) the users at S.José are moving with more viscosity (compare p. ex.
Radiology in S. José and Cardiology in S. Maria). For Cova da Beira we also know that
the population has a lower level of education. The numbers confirms an obvious conclu-
sion: Santa Maria does not have a sign system. Although, structurally very simple the
building without signs promotes confusion in way finding. Analysing all the results we
466 E. Côrte-Real, A. Ribeiro, S. André, S. Batel, C. Lisboa, C. Rosa

Table 22: Values for Cova da Beira Hospital

Medical services mrel Error Connectivity Integration2 vrel Viscosity
Medicine 6,19 3,37 49 92,16 14,54 ,98
Orthopaedics 6,19 3,37 50 96,40 14,84 ,95
Urology 6,19 3,37 50 96,40 14,84 ,95
Ophthalmology 6,19 3,37 88 114,49 26,11 1,41
Small surgery 6,19 3,37 89 84,64 26,41 1,93
Surgery 6,19 3,37 46 84,64 13,65 1,00
Paediatrics 6,19 3,37 49 64,00 14,54 1,41
Radiology 6,19 4,00 65 61,94 16,25 1,62
Day hospital 6,19 4,00 42 64,00 10,50 1,02
Special examine. 6,19 3,76 61 139,24 16,22 ,72
Blood harvest 6,19 3,76 39 34,22 10,37 1,88
Waiting room 6,19 3,76 8 16,00 2,13 ,82
Obstetrics 6,19 2,92 42 67,24 14,38 1,32
Physiotherapy 6,19 3,57 44 56,25 12,32 1,36

Table 23: Values for São José Hospital

Medical services mrel Error Connectivity Integration2 vrel Viscosity
Ophthalmology 6,19 4,65 15 9,00 3,23 2,22
Paediatrics Oph. 6,19 4,29 15 9,00 3,50 2,40
External Consult. 6,19 4,51 20 8,12 4,43 3,38
Radiology 6,19 4,88 48 23,04 9,84 2,64
Dentistry 6,19 4,20 19 11,56 4,52 2,42

Table 24: Values for Santa Maria Hospital

Medical services mrel Error Connectivity Integration2 vrel Viscosity
Surgery 5,85 5,50 42 67,40 7,64 ,66
Medicine 5,85 4,88 44 77,44 9,02 ,68
Orthopaedics 5,85 4,64 43 38,44 9,27 1,41
Otolaryngology 5,85 4,00 74 61,00 18,50 1,77
Dermatology 5,85 5,31 43 38,44 8,10 1,23
Corridor 5,85 4,36 26 75,00 5,96 ,47
Cardiology 5,85 7,00 48 24,01 6,86 1,67
Ophthalmology 5,85 7,00 57 46,24 8,14 1,03
 Pre-design of signage and orientation systems 467

Figure 186:

Table 25: Values for Santa Maria Hospital after the intervention
Medical services mrel Error Connectivity Integration2 vrel Viscosity
Medicine 5,9 6,32 44 77,44 6,96 ,53
Orthopaedics 5,9 6,00 43 38,44 7,17 1,09
Cardiology 5,9 3,41 48 24,01 14,08 3,43

also conclude that our formula exposes the malfunctions in the most integrated paths.
By using integration as “d” we can spot the low results that the system is performing on
those areas. In fact, we notice that Corridor, in Santa Maria, with high integration and
low “connectivity” should present values of Error near to 1 or 2 but still have relatively
high level of error. This result is enhanced by the mathematical proprieties of the formula
that shows a very low level of viscosity.
In order to test this model we decided to make an intervention in Santa Maria by
placing information signs on three paths: one with low viscosity (medicine), one with
medium viscosity (orthopaedics) and another one with high viscosity (cardiology). (Fig.
187) The signs were placed according to isovists and included pictograms since we had
found that when anxiety increases the ability of interpreting visual information increases
whereas the written information interpretation decreases. (Fig. 186-Fig. 188) The signs
were placed and after one month we conducted an inquiry on the exact procedures of the
previous one.
As we can see, viscosity was dramatically increased on the lower integration path and
surprisingly decreased on the other paths. Placing information on those high-populated
routes seemed to have increased confusion. Being more segregated, Cardiology, benefited
from the insertion of information. Going through areas of low population, people must
rely on signalisation. This was a pretty obvious conclusion that validates our formula but
an intriguing question remains about the other results that require further studies.

5. Main Findings

The main originality of this work is to focus on an idea of comfort in way finding. By mak-
ing the queries at destination points we are not measuring the real behaviour of fluxes but
468 E. Côrte-Real, A. Ribeiro, S. André, S. Batel, C. Lisboa, C. Rosa

Figure 187:

Figure 188:
 Pre-design of signage and orientation systems 469

how people qualify their movement. By crossing space analytical quantifiable information
(integration and connectivity) with cognitive ergonomics (interpretation of information)
and social psychology data (error) we also propose a way of enhancing the Space Syntax
methodologies. Although we state that the result of connectivity and integration should
be used directly as we did on the viscosity formula, both mass and time (interpretation
and error) can be computed with other data but of the same type.
As for what the universe of users analyzed is concerned we found no significant dif-
ferences between any of the hospitals, being the population of users constituted mostly
by women, elder, and low education level users as expected due to the universe analyzed
features. The existence of a “hospital culture” in Portugal, with a characteristic predomi-
nance of users that went there for more than five times (chronicle patients) was established.
Curiously, difficulties in way finding exist in most users, regardless of their previous usage
of the hospital. Thus, we believe that the introduction of signalization will affect both new
and common users significantly.
The lack of visual signals in the hospitals analysed, is the most probable cause for
the difficulty in way finding here observed in the queries, since from these one concludes
that when anxiety increases, pictorial information is more easily interpreted than writing
information. Since, in all cases, most people were in a relatively high level of anxiety one
concludes that the best way to improve flow is using, preferably, pictorial signage. The
only significant difference between all hospitals is the high percentage of first time users of
specific paths in Cova da Beira Hospital. The inexistence of a Health Care Center nearby,
responsible for a previous triage of patients, was by us identified as the cause.
Viscosity has shown itself rather regular in each hospital service. In a system, the
change of one parameter effects can be obtained from viscosity variation computation,
that is, viscosity, as we have defined it, can be use as an indicator for changes in the
system. Once related to signage systems placement, to a certain extent, we can test various
options by measuring the resulting viscosities in buildings and comparing with previous
values.
When a signage system is absent, a surprisingly high viscosity still characterizes fluxes.
We think that, on comfort in way finding, this kind of flux tends to be self-regulated due
to the expectations of difficulties related to real difficulties.
The high viscosity in S. José is due to the low connectivity in most of the paths
and probably due to the cluster organization which facilitates the creation of individual
cognitive maps. This would determine lower level of “error” than in the other Hospitals.
Trough the placement of signage in Santa Maria according to information theory principles
analysis and according to isovists methodology in three different paths, we measured
comfort in route finding by the same methodology here described, for both target paths
and unmodified paths in order to identify changes in viscosity. (Velichkovsky, B. M.: 1992)
(Souza, Carlos L.,1995) (Kuipers, B. J., 1983)
The queries, although depending of the quantity of people interviewed, already allow
a characterization of the flux inside hospitals structures. The characterization of speed
and mass can be more developed and accurate once accepted the principle of what kind
of information can be included and computed. For different built structures we must
respect the same principles. Also the inquiries must be performed in the same way after
introducing information in the systems.
Viscosity, since relates agents difficulties with integration is a correct measure of fluxes
distribution inside the buildings before and after the alteration made in the sign system.
It was also possible to verify that anxiety is a crucial factor of sign reading capability.
470 E. Côrte-Real, A. Ribeiro, S. André, S. Batel, C. Lisboa, C. Rosa

Figure 189:

It clearly shows that written and visual information must be combined to improve, for all
levels of anxiety, efficiency.
The combined methodologies, related through a formula, acts as a model tester. In
fact, confirming intuitive assertions about a built structure seems to verify the validity
of that formula. Certainly, the size of the data gathered will allow us to pursuit further
interpretations of the present project as will, also, be the basis for further projects. The
challenge will, as first further task, to relate the amount of information present in a system
with the changes in viscosity since we notice both increase and decrease. Another challenge
will be to compute architectonical information as friction on an inverted mathematical use
(more information=less friction).
For the time being we can, using average values for mass and error, anticipate, in
project, how a building will perform in terms of viscosity. We can also, using the “worse”
values, anticipate problems. By relating this study with studies on other kind of buildings
we may determine if the hospital users are the “worst” kind of users, in terms of comfort
in way finding and, therefore, use them as a tester for other projects. Using connectivity
and integration as the only variables on the formula will allow comparative studies of
different project options.
Trough this research project, we understood how “users” get comfortable while they
are walking from A to B, no matter if they are there by the first time, or if they are already
familiar with the hospital’s space. We translated space, through the gains/increasing values
of this multi-methodological approach, to the language of hospital users. The final goal,
to create a coherent sign system for all Portuguese Hospitals through a book of norms
is yet not concluded due to the lack of interest shown by the responsible people for the
country’s wealth care system.

1: http : //www.if.uf rj.br/teaching/f is2/hidrodinamica/viscosidade.html

2: http : //wejosephson.home.mindspring.com/tuskegee/visc.htm
3: http : //www.if.uf rj.br/teaching/f is2/hidrodinamica/viscosidade.html
4: http : //www.plus2physics.com/measurements/studym aterial.asp?chapter =
1&page = 2
Broadbent, G.; Bunt, R.; Llorens, T., (1980) Meaning and Behaviours in the
Built Environment, Chichester Wiley c.
 Pre-design of signage and orientation systems 471

Brown, D., (1998) Function Emergence, www.arch.su.edu.au/ ∼

scott/AID98/emergence − workshop/pa.../brown.htm
Carpman, J. & Grant, M. [(2002) Wayfinding: a broadview. In: R. B. Bechtel & A.
Churchman (Eds.), Handbook of Environmental Psychology, NY: John Wiley & Sons,
Inc.
Carpman, J.; Grant, M., (1993) Design that Cares - Planning Health Facilities for
Patients and Visitors, American Hospital Publishing.
Carvalho Rodrigues, F.; et. al., (2001) Quantidade de Informação na Arquitec-
tura Portuguesa, Comunication, presented at the Lisbon Academy of Science.
Cherulnik, P., (1993) Application of Environment - Behavior Research, Case Studies
and Analysis, Cambridge University Press.
Costa, J., (1992) Imagem Publica- Una Ingenierı́a Social, Madrid, Fundesco.
Costa, J., (1971) La Image y el impacto Psico-Visual, Barcelona, Ediciones Zeus.
Costa, J., (1989) Siñalética, Barcelona, Ediciones CEAC, S.A.
de Long, A. J. Coding Behaviour and Levels of Cultural Integration: synchronic and
diachronic adaptive mechanisms in human behaviours, in: Meaning and Behaviour
in the Built Environment.
Doise, W.; Deschamps, J.; Mughy, G., (1980) Psicologia Social Experimental,
Tradução Adelino Gomes, Lisboa, Morais.
Durso, F.(ED), (1999) Handbook of a Applied Cognition, New York, John Wiley &
Sons.
Ferreira, V., (2001) Mapas Axiais, Espaços Convexos e Isovistas, A Teoria da Sintaxe
Espacial e o Uso de Aplicações Informáticas, Lisbon, Policop, FAUTL.
Gero, J., (1998,) Emergence in Design, Lisbon, AID Workshop 5. Functional Emer-
gence.
Hillier, B., (1998) Space is the Machine, Cambridge University Press.
Hillier, B. and Hanson, J., (1984) The Social Logic of Space, Cambridge University
Press.
Kuipers, B. J. (1983) The Cognitive Map Could it have been any other way? Spatial
Orientation: Theory, research and application, Herbert L. Pick, Jr and Linda Peds.
Acredolo New York: Plenum, 1983, p. 345 - 359.
Lima, M., (1998) Factores Sociais na Percepção de Riscos, in: Psicologia XII (1), p.
63-84.
Peponis, J., (2001) Interactions Questions and Descriptions - How do they look from
here?, in: Proceedings Space Syntax. 3rd International Symposium Peponis J (ed.),
Georgia Tech, Atlanta.
Simões, A.; Carvalhais, J.; Trindade, C., (1998) New Information Systems on
Public Transport: A way of increasing mobility of erderly an disable people, Con-
ference Proceeding of the 8th International Conference on Transport and Mobility
for Elderly and Disable People (TRANSED’98), Perth, Western Australia, Vol. 2,
p. 839-847.
Souza, Carlos L. (1995) Cognição Ambiental e as relações: Mapas Cognitivos, Ambi-
ente Construı́do & APO, textos do Laboratório de Psicologia Ambiental, Vol. 4, n 8,
Instituto de Psicologia da Universidade de Brası́lia.
472 E. Côrte-Real, A. Ribeiro, S. André, S. Batel, C. Lisboa, C. Rosa

Smyth, M.; Collins, A.; Morris, P.; Levy, P., (1994) Cognition in Action,
Lawrence Erlbaum Associates Publishers.
Sudman, S.; Bradburn, N.M., (1996) Thinking about application of cognitive pro-
cesses to survive methodology, San Francisco, S.A., Jossey- bass.
Velichkovsky, B. M. (1992) The Spatial representation system: a single system of
perceptual - verbal access? PSYCOLOQUY 3(46) space 7.