Wayfinding mindmap

Cognitive maps

The general consensus is that we use a representation called a cognitive map to find our way around. These maps serve to represent places, spatial relationships and travel plans. The term cognitive map was first used by Tolman (1948). He deduced that rats that could find their way around (and out of) a maze without having been subjected to reinforcement or reward schedules must have some type of mental structure which they had "built" mainly from observation, but also from some limited experience.

Some recent research by Pearce (2007) elaborates on some of the complexities of these issues in rats and other animals and suggests that the mental representation that animals have of the world does not involve a full knowledge of spatial relationships and that animals are incapable of representing such symbolic or abstract thoughts. You can read his paper "How does a yak find a drink?" here.

Lynch (1960) asked residents of Boston, Los Angeles, and Jersey City to draw sketch maps of their respective cities showing the routes that they commonly used. This classic study found that people universally produced drawings with:

Appleyard (1970) asked people to produce sketch maps of Venezuela. He claimed two distinct types of map:

  1. The serial or sequential map  where major points are indicated in an ordinal relationship way - a bit like the London underground map.
  2. Birds eye-view  maps showing an overall pictorial view, demonstrating survey knowledge.  

Familiar places tended to be more accurately mapped and when asked to draw larger areas than just their locality, people tended to start with the familiar area in the centre, and using this point as an anchor, moved outwards. It was also found that people tended to expand the size and detail of familiar areas.  

A link to some examples of University of Northampton student maps can be found here. Students were asked to draw a map of the route from the University to the railway station in Northampton. Male students were generally more accurate and their maps had more detail. Female maps showed more use of landmarks and less accuracy concerning roads. Interestingly the male maps were drawn from a "map perspective" with the University at the top of the page (north). Female maps were drawn from a personal perspective with the town going away from their own fixed point of view. This is a typical result from this sort of exercise

There are parallels here with other phenomena that we know of. When people are asked to make size and distance estimates they tend to use themselves as an anchor or reference point in space. Also in the history of cartography (real map drawing) the centralization of important local features is (almost universally) found, and peripheral areas are vague or get imputed with boundary or other features. In many early mappae mundi (maps of the world), usually drawn by monkish scholars, Jerusalem was placed at the centre of the world and peripheral features were vague and labelled "here be dragons" or such like. If you like maps and things have a look at chapter 1 in Dorling and Fairbairn (1997) Mapping - Ways of representing the World – classmark 304.2 (in the Geography section!) This shows very interestingly how the nature of maps and representations of the world tend to mirror, the culture and thoughts and understanding of the people that composed them.

Follow this link to a study by Moar 1978 which demonstrates this.

The primary function of a cognitive map is for wayfinding. So cognitive maps have an adaptive function and must in evolutionary terms have served to allow us to find food, shelter, meet with others etc in an efficient way. Cognitive maps enable us to plan for action and also to follow directions and instructions. And there is the implication that cognitive maps can be the basis for communication, if we assume that we all share approximately the same sort of representation.

Garling (1984) suggests that wayfinding requires a sense/understanding of destination, location, route, and choice of mode of getting from location to destination. After a journey information is fed back into the system to improve efficiency of the next trip.

Some people such as Passini (1990) suggest that wayfinding requires no overall pictorial representational map, but is merely a sequence of decisions based on recognising the correct choice at each point along the route.

I guess these two ideas and the finding of Appleyard (above) exemplify one of the main dichotomies in wayfinding. One school suggests a holistic approach and the other a more sequential, propositional view of the process. This may be a false problem, as Tye suggests that we use both representational systems, but we use the propositional system to (re?)generate a holistic representation.

Things are further complicated by the relative legibility of the environment we find ourselves in.

A legible environment

Some places e.g. cities with an easy to see "Centre" are easier to find your way in than others which may lack focus.

Garling suggests the following four influences of the environment on legibility and wayfinding:

  1. Differentiation. High differentiation environments look different. Low differentiation ones look very similar. An environment which has buildings that stand alone or are unusual in shape are better remembered than others.
  2. Degree of visual access. Can the setting be seen from different angles. Those that can are more legible and give rise to more accurate cognitive maps.  
  3. Complexity. Information needed to get around e.g. floor plans in buildings etc can be simple or complex (or at least presented more or less simply) Complex environments are less legible.  
  4. Transitions. Places which have high numbers of transition points make orientation easier. e.g. multiple exits from buildings onto a variety of different streets supposedly enhances sense of direction. (Although I guess there’s probably a trade-off here, too many exits will lead to greater complexity and this should be detrimental to wayfinding.)

 

Important factors in cognitive map formation

Time and experience. Studies looking at the development of cognitive maps in students have shown that these improve with exposure, maps are more accurate representations after a year on campus.  

Nature of exposure/activity. Drivers seem to have better organised spatial information systems than passengers.  

Contrasts can be seen between urban and rural people. Those used to living in urban environments use different cues or at least have salience differences when compared to country dwellers e.g. country people may make distinctions between say a hazel tree and an ash tree, whereas urban dwellers may see all green things on sticks as just trees. City dwellers make distinctions between various urban structures that country dwellers miss.  

Orleans (1973) suggests that rich and poor have differing cognitive maps of similar areas. This is probably a specific case of the time and experience phenomenon – poor people have less mobility, restricted to public transport etc and don’t build up extensive or wide-ranging experience.  

Appleyard (1990) suggests that male maps are generally more accurate than female ones. Males supposedly emphasise routes, whereas females are supposed to emphasise landmarks more. Males giving directions are more likely to give compass bearings and distances and make fewer errors.  Compass directions may be odd to us Brits, but I was watching an old black and white gangster movie the other night and the evil kidnapper gave the hero the following directions over the phone. "Ya go out tha lobby of the hotel. Go three blocks north on twenty third street . . . . " – a cultural phenomenon perhaps?

Orleans (1972) suggests that females are more home-centred. Women’s maps tend to be more detailed for areas closer to their homes, and home neighbourhoods (remember the distorted cognitive map of Scotland from lecture 2 – that map came from women and showed a similar elaboration about "home territory"). Men are supposedly better at mapping wider ranges of surrounding areas. Men also learn better than women when exposed to an area by being driven around it. But both sexes are supposedly equal when it comes to learning the way around a new environment on foot. Bryan suggests that these differences are real, but the issue is confused by the fact that "a good sense of direction" is more important to the self-esteem of men than women, so it’s more important for a man and so he takes more notice of the environment etc. On the other hand it might be that men are better than women and that women increase the differential by relying on men to find the way for them and men comply. Anyway—it’s a thorny old issue—there are loads of old studies which suggest that women have worse spatial abilities than men, and there are clearly strong socio-cultural effects at work here. Most of these studies were done on Americans and most of the studies are a good few years old now. It might be the case that these sorts of findings are very dependent on time, place and culture. I suppose lots of environmental psychologists would say that they ought to be, because the effects of environment are highly dependent on the individual features of the situation, people involved, purpose etc. And now, with the increasing availability of satellite navigation systems, perhaps we may lose our ability to find our way around the world completely.

Passini (1990) in an interesting study says that there are no (very few) differences between the wayfinding abilities of blind and sighted people. The suggestion is that blind people rely more on sequences of propositional statements rather than on a "mental image" type of map. As well, blind people tend to make greater use of tactile cues than sighted people.

For an interesting account of wayfinding in the blind, follow this link to a study by Simon Ungar.


How is Cognitive Map information stored?

Clearly maps can be either analogue, "pictorial image" entities, or lists of propositions or statements that are interconnected. Tye (1991) reckons that both sorts of representation are used. Probably propositions are used for long term storage of the information, and that these are used to (re)construct an analogue map for short term usage. Certainly it must be the case that if people can accurately (or even not-so-accurately) produce estimates of distances, there must be some propositional storage. But there is strong evidence (Baum 1977) that analogue systems are used widely. If a person is required to consult information about large areas or situations where there are many areas to be scanned, then the person takes longer to give correct answers. If a person has many distances to estimate (so they are "adding the distances" in their heads) then that will take longer the longer the distances that are actually involved. Propositional storage should allow that information to be accessible more quickly than is the case in reality. For an analogy have a look at cognitive psychology studies – there’s one by Shepherd and Metzler where people have to do mental rotation tasks, and the greater the amount of actual rotation that a person has to do, then the longer reaction time it takes to do it. This seems to suggest that there is a real pictorial image that is somehow being manipulated.

Allen suggests that it is possible to see "chunking" of spatial information in cognitive maps, in much the same way that chunking increases the capacity of short-term memory. Allen suggests that we chunk areas together in clusters, marking each cluster with a landmark. Work by Holding (1992) shows that where there are clusters with more than one landmark distances between landmarks are judged to be closer together than those which are in different clusters, even if the objective distance is the same.

Errors in cognitive maps

Are not random but show four characteristic features:

  1. Incompleteness. Omissions are usually of minor pathways and details, but sometimes complete districts and landmarks can be left out.
  2. Augmentations. Including features which are not there. This is relatively rare.
  3. Distortions. Usually landmarks are too close together or too far apart, or are in incorrect alignment. Lines are often made parallel when they are not so in reality. Junctions are frequently made to be right angles when they are not. De Jonge reckons that people remember straight lines better than curves and right angles better than any others. Does this suggest that urban planning ought to take geometrically (gridiron) based cities etc as a better starting point? It is often noted that places with angular features are less easy to get lost in than others. But interestingly (?) most studies of environmental aesthetics seem to say that people prefer gentle curves and meanders to straight lines and right angles. If people do distort curves into straights and other angles to right angles - what might that suggest about the analogue/propositional debate? Might it tie in with some of the visual brain feature-detector types of studies (Barlow and Blakemore etc)? Further thoughts on brain and space below.
  4. Superordinate scale bias. This is an extension of work by Tversky which highlights the illogicality of human thinking and reasoning, and how biases seem to systematically distort our cognition. The superordinate scale bias in environmental psychology shows itself by knowledge of larger areas distorting our knowledge for smaller areas. So our global knowledge (or lack of it) may distort other judgements. e.g. we know that Edinburgh is on the east coast of Scotland. But if you consult an atlas it is actually further west than Bristol which we know is on the west coast of England! And of course there is the differential perception of understanding of N/S E/W spaces anyway which may confuse things even further. Another related (?) finding is that people are better at making N/S judgements than they are at making E/W ones. Most folk will say that Sheffield is north of Birmingham, but who can say whether Sheffield is East or West of Birmingham?

A physiological basis for cognitive maps?

There are certainly some studies that suggest that cognitive maps may be located in the hippocampus. Some neurons there are thought to be coded for place, those in the left hemisphere storing/sorting the verbal components and the right hemisphere housing the pictorial map. Damage to the parietal lobe of the cortex can lead to the loss of the ability to draw maps and loss of knowledge about how places are arranged geographically. Does this help or hinder the analogue/propositional debate?

Taxi driver brain study

A study by Maguire looking at the brains of London taxi drivers seems to show that the hippocampi of people who have had extensive route training and who possess extensive route knowledge of a city, become enlarged. And another study by the same author would seem to show that this training can generalize to new environments where the taxi drivers seem to be able to absorb wayfinding information better than control participants.

How to improve wayfinding in towns and cities

Some environmental psychologists ever wanting to improve the lot of humankind, have suggested that cognitive mapping features should be used to provide better legibility of the environment and hence better wayfinding advice for people in strange (novel) environments

So to improve wayfinding in cities:

  1. Landmarks should be placed at major decision points in the road system. Urban/road planners in Britain have a penchant for putting featureless roundabouts in such places. Having special features on each should make wayfinding easier.
  2. Landmarks of whatever sort should be highly visible, tall and distinctive (unique in style). And also, when landmarks compose a strong feature of a community’s cognitive map, don’t do daft things like rename the pub on the corner. Those who’ve lived in Northampton long enough will remember the saga of the White Elephant pub. This well-known landmark on the corner of the Racecourse was refurbished to become the Kingsley Park Tavern. But locals insisted on calling it by its old name, buses stopped at the White Elephant, and if you were a stranger to the area and asked directions from people you'd end up getting very confused very easily. Sanity eventually prevailed and the White Elephant is now the White Elephant again.
  3. Road signs and names should be meaningful and distinctive. Again there is a disturbing trend to name housing estates around a certain theme, where all the names are similar. Also these developments also tend to have mass-built housing which tends to look uniform and lacks distinctiveness. An example of how not to name roads.
  4. Make road systems correspond with district boundaries.
  5. Preserve buildings that make good landmarks, and construct landmarks in areas which lack them. Features presented to new wayfinders like "Here you are" maps should always be in the current orientation for the observers position, include a direction arrow to show which way the observer should be looking, and should also have labels that match the environmental features. If you haven’t been to Brighton, go and visit the Lanes. - A dreadful little warren of shops and streets which has in its centre a map which says "You are in the Lanes" and gives almost no further information. It is also presented on a horizontal surface in such a way that true N/S is aligned NE/SW. However there are tourist maps of the area which show pictograms of major buildings and roads and are better in terms of corresponding with the cognitive maps of people, and which people find easy to use. Many other tourist spots have adopted this type of approach, and surveys of users tend to be supportive of their use. The Open University have been particularly successful in using pictographic maps to enhance wayfinding

Chapman (1984) suggests that the average person’s wayfinding ability in most British towns could be improved by about 35 to 40% if all shops and buildings were clearly numbered. On Saturday, have a wander down Abington Street, & Gold street in Northampton and see what our own dear little town is like.


Development of wayfinding ability.

Most studies of cognitive mapping look at those of adults, possibly because drawing sketch maps etc is really a task better suited to adults than to children. But there is clear evidence that we develop our skills at wayfinding from childhood until we are adults. Clearly developmental theorists such as Piaget have much to say about this - the interaction with objects is one of the key underlying conceptions of Piagetian theory. As with most of Piaget’s stage theorising there is a suggestion that wayfinding behaviour shifts to a more abstract process from a reliance on knowledge of the body and it’s immediate surroundings. Also it may be that different types of learning affect the way in which we develop awareness of space and place.