Lynch Maps Lie When Grain Is Invisible: Why Indian Sites Need Image, Morphology and Configuration Read Together
Urban design · 10 Sep 2026 · revised 2 Oct 2026
Lynch Maps Lie When Grain Is Invisible: Why Indian Sites Need Image, Morphology and Configuration Read Together
Kevin Lynch's five elements are still the best way to teach legibility. On Indian streets, though, an image drawn without grain, and landmarks drawn without the network that feeds them, give you handsome boards that miss how people actually move, stop and claim a place. This essay puts Lynch, Conzenian morphology, Space Syntax and the public-life tradition into one diagnostic you can run on a real corridor.
Walk out of almost any Indian railway station toward the old bazaar and you can feel the mental map redraw itself every few hundred metres. The forecourt is a node with no room to stand in. The rail yard or the flyover is an edge that cuts rather than defines. The temple spire you were steering by shows for three blocks, then disappears behind hoardings and setbacks. The lanes of the old core make no sense on a tourist map, yet the vegetable seller moves through them without a second thought, because she knows every otla, every gate and every short cut.
Faced with this, most consultants draw a Lynch diagram: paths, edges, districts, nodes and landmarks in five colours. The board looks finished. The diagnosis has not started.
Lynch never said imageability was the whole of urban design. The Image of the City was an empirical study of how people build a mental picture of a city from five kinds of physical element (Lynch, 1960). Those five categories are still indispensable. What has moved on is the evidence around them. We now have good reason to believe that three other things decide whether an image holds or falls apart: the grain of the built fabric (figure-ground, plots and blocks), the configuration of the street network (how integrated a street is, how much through-movement it carries), and the depth of the edge where building meets street. Use Lynch as a stand-alone checklist and you get scenic legibility. Read him together with morphology and configuration and you get a brief.
So here is the argument, stated plainly for Indian practice. A Lynch map that leaves out grain and network configuration is not neutral. It misreads organic fabric, overrates monumental landmarks, and says too little about the public realm, which is where a place's identity is actually lived. The fix is not to drop Lynch. It is to read the site in layers before redrawing anything.
Three ways of reading, one city
1. Image: the city people carry in their heads
Lynch's elements remain the clearest words we have for structure as people perceive it. Paths matter because we experience a city in sequence, on the move. Edges mark limits, and sometimes they block. Districts hold a character at a scale you can feel yourself entering. Nodes are where choices concentrate and people pause. Landmarks give a fix from outside (Lynch, 1960). A place is imageable when these elements support one another; one icon cannot do it alone.
Two habits go wrong in practice. The first is drawing landmarks too heavily, because they photograph well and are easy to locate. Districts and edges, which in Indian cities often carry the cultural meaning, get a dotted line and little else. The second is treating paths as traffic capacity rather than as a sequence of experiences. Carmona's framework of urban design dimensions is a useful corrective: place-making is continuous work across form, perception, social use and process, not a single exercise in branding icons (Carmona, 2021). Lynch is where perception starts. He does not tell you about morphology or movement.
[[FI
G01]]
2. Morphology: what the solids, plots and blocks really are
Urban morphology supplies the grammar of solid and void that a Lynch map lacks. Conzen's town-plan analysis separates three things: streets and their system, plots and the blocks they make, and the footprints of buildings. Together they form a durable frame inside which uses and façades come and go (Conzen, 1960; Whitehand, 2001). The figure-ground drawing makes that frame visible in a second. Rowe and Koetter's reading of the city as collage taught studios to see the void as a figure in its own right (Rowe & Koetter, 1978). Trancik gave us the triad of figure-ground, linkage and place, and the useful phrase "lost space" for the leftover ground around object buildings (Trancik, 1986).
Grain is not an aesthetic preference; it changes how a street works. Fine grain, meaning small plots, many doors and short blocks, gives you many ways through and many thresholds. Coarse grain, meaning big footprints, blank plinths and superblocks, gives you long detours, weak character and edges that act as walls rather than seams. Many Indian old cores have a dense, continuous figure-ground, laced with courts and galis that a Nolli-style reading would show as public or semi-public ground. Much of the post-1980s growth along corridors and around stations is the reverse: object buildings standing in residual space. Draw one Lynch "district" over both without a grain map and you will treat them as the same kind of place. They are not.
Pune makes the point at one scale. Along the walk from the station to the Mandai, a 400 m window in the peths has 44% ground coverage, 34 buildings per hectare and a median footprint of 90 m², as mapped in OpenStreetMap. A window of the same size in the station quarter has 19% coverage, 3 buildings per hectare and a median footprint of 490 m². More than a quarter of that second window lies farther than 20 m from any wall: space no building gives shape to. In the peth window the share is effectively zero.
[[FI
G02]]
Grain also lives in the street network. Count how often the public streets let you turn and the same two fabrics separate again: 300 junctions per square kilometre in the peth window against 69 in the station quarter, and a median block perimeter of 280 m against 550 m. Going round a block in the station quarter takes roughly twice the walk it takes in the peths.
[[FI
G03]]
3. Configuration: why some streets become destinations
Space Syntax, built up by Hillier, Hanson and colleagues at UCL's Bartlett, treats urban space as a system of relations: where each street sits in the whole network shapes how much movement it is likely to get (Hillier & Hanson, 1984; Hillier, 1996). The claim that matters most to designers is natural movement. The street grid itself is a primary driver of where people walk; shops and other attractors then multiply those flows rather than create them from nothing (Hillier et al., 1993). Integration measures how close a street is, in turns, to every other street. Choice measures how often it lies on the simplest routes between other places. Between them they predict reasonably well where people will walk, and therefore where active ground floors, stalls and informal resting places tend to survive. For a compact glossary of the measures and their limits, Yamu, van Nes and Garau (2021) is the most useful single source.
None of this means culture does not matter. Raman and Dempsey's study of Ahmedabad's walled city found that Muslim neighbourhoods there have the integrated public spaces clustered at their centres, much like settlements that grew without a plan, while Hindu neighbourhoods are "inside-out", with their most integrated spaces at the edge. The two forms shape how communities meet and keep apart (Raman & Dempsey, 2012). Configuration is a lens, not a fate. But if you redesign a bazaar street, a station approach or a heritage fringe without it, you are rearranging the shops without understanding the movement that keeps them alive.
[[FI
G04]]
4. Thresholds and human scale: where an image becomes a place
Between the building and the street lies the edge where public life happens or fails to. From Life Between Buildings to Cities for People, Gehl's position has been consistent: the activities people choose to do, and the social ones that follow, depend on soft edges. That means frequent doors, things to see, niches, places to sit, shade, and distances matched to the human senses (Gehl, 2010, 2011). Hard edges, such as blank walls, parking in the setback and dead plinths, carry people past but never invite them to stay. Thwaites, Simpson and Simkins frame the street edge as a "transitional edge" with social, spatial and material dimensions, and they read its properties along the edge, across it and within it (Thwaites et al., 2020).
In Indian practice this is the layer that decides the outcome. The otla, the arcade, the shop's spill onto the pavement, the temple forecourt and the mouth of a gali are not quaint vernacular. They are the depth where Lynch's path, Gehl's edge and the morphologist's plot frontage meet. Take them away for road widening or "façade improvement" and the landmark's silhouette may survive while the district's image and the node's purpose do not.
[[FI
G05]]
5. Evidence: closing the loop
The evidence-based design literature now argues for a working method that is iterative and moves between scales, testing proposals against data at each step. Karimi's earlier paper showed how the configurational approach becomes an analytical design tool (Karimi, 2012). He and his colleagues have since set out a full framework for evidence-based design and planning, drawn from six projects that range from public-space upgrades to metropolitan plans (Karimi et al., 2025). The modelling has moved too. Ullrich and colleagues couple a city-scale network model with an agent-based model of walkers and test it against counts in central Gothenburg (Ullrich et al., 2024). Ortner and colleagues generate thousands of neighbourhood layouts and score each against a set of walking-accessibility measures, which exposes the trade-off between economical and pedestrian-friendly plans (Ortner et al., 2025). A systematic review of space syntax studies of sociability also maps where the evidence is thin (Askarizad et al., 2024).
The lesson for practice is humility about method. Each layer of theory should produce a hypothesis that fieldwork and simple analysis can prove wrong.
The IGIT Diagnostic: Image, Grain, Integration, Threshold
What follows is a stepwise protocol for reading a site before concept design. I call it the IGIT Diagnostic, after its four layers. It is built to work on an Indian corridor or old core with ordinary base maps, a day or two of fieldwork and, where you have it, Depthmap or a similar network tool. It does not replace the statutory process. Its job is to discipline the designer's reading so that the proposal answers the right failure.
[[FI
G06]]
Step 0: Bound the study and state the question
Draw a walkable study area, usually 800 m to 1.5 km around a station, a chowk or a heritage gate, and write down one question. "Why does the walk from the station to the bazaar feel broken?" "Why is the heritage district losing people who stay, even though footfall is high?" Without a question, Lynch mapping turns into an inventory.
Step 1: Image layer (Lynch structure)
Map the paths, edges, districts, nodes and landmarks from desk study and from mental-map interviews. Eight to twelve regular users is enough to start: vendors, students, auto drivers, residents. Record the conflicts: edges that block, nodes with nowhere to stand, landmarks you can see but not reach, districts that change character with no readable boundary. The output is an annotated Lynch map and a short conflict list. Do not propose anything yet.
Step 2: Grain layer (morphology)
Draw a figure-ground at one scale for the whole study area, and classify blocks and plots by grain: fine, medium or coarse; continuous or broken. Look for the Conzenian clues where they apply: older street and plot patterns that still constrain new building, fringe belts along the railway or the river, and plots merged until their thresholds vanished. Lay the grain over the Lynch districts and ask whether each "district" shares one grain or whether the label is hiding fabrics that do not belong together. The output is a figure-ground and a grain typology map.
Step 3: Integration layer (configuration)
Build or obtain a segment or axial model of the network people can walk. Include the galis that work as through-routes; leave out private courts unless the public actually uses them. Compute or estimate global and local integration and note the high-choice streets. Check them against pedestrian counts at six to eight gates. Where high integration meets a blank edge, you have a movement-interface mismatch. Where a major cultural landmark sits on a poorly integrated street, you have an access-meaning mismatch. The output is an integration map and a mismatch table.
Step 4: Threshold layer (the public-realm edge)
Survey sections and frontages along the critical paths, in samples of 200 to 400 m. Count doors, and record active against blank plinth, shade, level changes, seating, informal spill-over and sightlines to landmarks. Classify each edge as soft, hard or transitional. Then map how rich the threshold is against integration. The design opportunity usually sits where high integration meets a hard edge, or where a fine soft edge is stranded on a dead end. The output is an edge typology plan and two or three analytical sections.
Step 5: Site lenses (climate, ecology, views, heritage)
Only now lay on the lenses that are not about form: microclimate (shade corridors, open spaces that overheat), ecology (the nala, continuity of tree canopy, ground that can flood), views (landmark corridors to protect or open), and heritage (listed buildings, living cultural routes, intangible patterns of use). These lenses filter what IGIT found. If they come first, you risk conservation or greening strategies that fight the movement economy.
Step 6: Synthesis brief (failure modes into moves)
Turn the mismatches into a short brief built on named failure modes, for example:
- Severed path: a remembered path cut by a morphological barrier or a break in the network.
- False node: a chowk with a name but no room, shade or soft edge.
- Landmark without approach: a visual landmark sitting on a poorly integrated street.
- Grain rupture: a coarse insertion tearing a fine district.
- Threshold starvation: high integration and a blank plinth.
Each failure points to a family of moves: reconnect, thicken the edge, repair the grain, give the node a room, re-sequence the path. Test each move again against integration and the climate lens before detailing.
Worked sketch: from a station to the bazaar core
Picture a composite mid-size Indian city. A colonial-era railway station sits a kilometre or so from a walled or dense bazaar core. A 1980s arterial has become a "corridor" of large retail footprints. A nala is both an edge and what is left of the ecology. A temple shikhara still orients the old town. The figures test this sketch against one real corridor, the walk from Pune station to the Mahatma Phule Mandai across the Nagzari nala, drawn from OpenStreetMap. The shortest public route is 3.2 km.
Image. Users draw a strong path along the arterial, because vehicles dominate it, and a weaker, broken walking path through service lanes. The nala reads as an edge, mostly a barrier. The station is a node without a public room. The bazaar is a district with intense character but a blurred outer boundary since the walls came down. The temple is the main landmark; a new mall tower competes for the skyline without helping anyone find their way on foot. In Pune, the walk passes under the rail yard through an underpass, and that is where the mental map breaks.
Grain. The bazaar figure-ground is fine and continuous, and its courts and galis form a second public network. The corridor figure-ground is coarse: islands of building in parking and setback, Trancik's lost space. Where the corridor meets the old town, plot amalgamation has already wiped out two frontages rich in otlas. The consultant's Lynch "district" boundary sits right on this grain rupture. IGIT would move the district edge to the morphological frame, not leave it on the ward line. Measured on the Pune walk, the station side has a junction every 93 m and the peth side one every 42 m. The station-quarter blocks have a median area of 1.47 ha against 0.41 ha in the peths.
Integration. Local integration stays high inside the bazaar's lane network, which fits a market that is still alive. Global integration favours the arterial, which carries through-movement but has almost no soft edge. The worst mismatch is the station approach: high choice for vehicles, a broken route for walkers, and the temple pushed to the margins of the station's natural movement. "Beautifying" the arterial for retail will fight the bazaar's movement economy unless the walking network is repaired first. On the Pune drawings, 45 stretches of the most integrated streets have a built edge along less than a fifth of their length, and 39 of them lie on the station side of the Nagzari.
Threshold. The bazaar frontages are soft: many doors, and shade from the upper floors that project over the street. The arterial is hard: glazed but set back, and exposed to the heat. The temple forecourt is a true transitional edge, under pressure from parking. Seen through the climate lens, staying on the arterial collapses on summer afternoons, and the canopy inside the core is the real comfort infrastructure. Along the Pune walk, the station half has no building within 14 m of the centreline on either side for 83% of its length; in the peth half the figure falls to 37%.
Brief. The failure modes are a severed walking path from station to core, a false node at the station, threshold starvation on the high-choice arterial, a grain rupture at the meeting of corridor and core, and a landmark whose approach the network ignores. The moves follow the walk, in order:
- Give the station forecourt a room: soft edges, a place to stand, and a clear line of sight and approach to the temple route.
- Run a continuous shaded walking spine through the residual voids and the underpass, following streets of medium to high local integration rather than forcing walkers onto the arterial kerb.
- Treat the nala as a seam, an edge that connects, rather than a fenced barrier, and make its crossing a named node.
- Stop further plot amalgamation at the grain rupture, and require active, frequent thresholds on any redevelopment that faces the repaired spine and the bazaar approach.
The landmarks are kept. The identity is rebuilt at the grain and at the threshold.
Conclusion
Lynch gave urban design a shared language for legibility. Morphology gave it the durable structure of streets, plots and buildings. Space Syntax gave it a way to predict how that structure distributes movement. Gehl and the work on transitional edges gave it the human edge where movement turns into place. Indian sites, with their dense old cores, improvised thresholds and abrupt corridor insertions, show what it costs to use any one of these alone.
The IGIT Diagnostic is not a new ideology. It insists on an order: image, to name the structure people carry in their heads; grain, to test whether that structure is real on the ground; integration, to find the movement economy and its mismatches; threshold, to specify the public realm; and then the site lenses, to filter for climate, ecology, views and heritage. Read the site in that order and your Lynch map stops lying. Leave grain and configuration out, and you will keep polishing landmarks while the street, where the city's image is actually made, slips away.
References
Askarizad, R., Lamíquiz Daudén, P. J., & Garau, C. (2024). The application of space syntax to enhance sociability in public urban spaces: A systematic review. ISPRS International Journal of Geo-Information, 13(7), 227. https://doi.org/10.3390/ijgi13070227
Carmona, M. (2021). Public places urban spaces: The dimensions of urban design (3rd ed.). Routledge.
Conzen, M. R. G. (1960). Alnwick, Northumberland: A study in town-plan analysis (Institute of British Geographers Publication No. 27). George Philip.
Gehl, J. (2010). Cities for people. Island Press.
Gehl, J. (2011). Life between buildings: Using public space (J. Koch, Trans.). Island Press. (Original work published 1971)
Hillier, B. (1996). Space is the machine: A configurational theory of architecture. Cambridge University Press.
Hillier, B., & Hanson, J. (1984). The social logic of space. Cambridge University Press.
Hillier, B., Penn, A., Hanson, J., Grajewski, T., & Xu, J. (1993). Natural movement: Or, configuration and attraction in urban pedestrian movement. Environment and Planning B: Planning and Design, 20(1), 29–66. https://doi.org/10.1068/b200029
Karimi, K. (2012). A configurational approach to analytical urban design: 'Space syntax' methodology. Urban Design International, 17(4), 297–318. https://doi.org/10.1057/udi.2012.19
Karimi, K., Zhand, S., Simons, G., Abdeldayem, W. S., Charalambous, N., & Giraud, I. (2025). Framing evidence-based design and planning: An analytical, multi-scalar and iterative framework for urban design and planning. Urban Science, 9(11), 457. https://doi.org/10.3390/urbansci9110457
Lynch, K. (1960). The image of the city. MIT Press.
Ortner, F. P., Chen, Z., Aydin, E. E., & Song, P. (2025). Generative urban modelling for walkable neighbourhoods: Design optimization for pedestrian-oriented street networks. Journal of Urban Design, 31(4), 515–534. https://doi.org/10.1080/13574809.2025.2517652
Raman, S., & Dempsey, N. (2012). Cultural diversity and spatial structure in the Indian urban context. Journal of Urban Design, 17(3), 425–447. https://doi.org/10.1080/13574809.2012.683399
Rowe, C., & Koetter, F. (1978). Collage city. MIT Press.
Thwaites, K., Simpson, J., & Simkins, I. (2020). Transitional edges: A conceptual framework for socio-spatial understanding of urban street edges. Urban Design International, 25(4), 295–309. https://doi.org/10.1057/s41289-020-00115-9
Trancik, R. (1986). Finding lost space: Theories of urban design. Van Nostrand Reinhold.
Ullrich, A., Hunger, F., Stavroulaki, I., Bilock, A., Jareteg, K., Tarakanov, Y., Gösta, A., Quist, J., Berghauser Pont, M., & Edelvik, F. (2024). A hybrid workflow connecting a network and an agent-based model for predictive pedestrian movement modelling. Frontiers in Built Environment, 10, 1447377. https://doi.org/10.3389/fbuil.2024.1447377
Whitehand, J. W. R. (2001). British urban morphology: The Conzenian tradition. Urban Morphology, 5(2), 103–109. https://doi.org/10.51347/jum.v5i2.3896
Yamu, C., van Nes, A., & Garau, C. (2021). Bill Hillier's legacy: Space syntax—A synopsis of basic concepts, measures, and empirical application. Sustainability, 13(6), 3394. https://doi.org/10.3390/su13063394
Figures: map data © OpenStreetMap contributors, ODbL. Measures were computed on OSM footprints and street centrelines as mapped in October 2026. Integration and choice are ranks from a simplified angular segment analysis, not Depthmap output. Sections are indicative, not surveyed.
Comments
Post a Comment