Invisible Maps of Trade

Compare the Carta Pisana with the anonymous fourteenth-century chart held by the Library of Congress. Their outlines are not identical, nor are their surviving colours, annotations or emphases. Yet they share a way of making the sea legible. The coast is a chain of names. The rhumb network offers a disciplined visual field. Promontories and islands punctuate a surface whose practical meaning lies in passage.

The shore is therefore more than a boundary between blue and brown. It is infrastructure.

Infrastructure is usually imagined as something built: mole, arsenal, warehouse, beacon, galley, road. On a portolan, part of that infrastructure is editorial. A place-name must be selected, repeated and placed where it can be read from the sea. A coastline must be stabilised through copying and correction. A crowded cluster must still allow the navigator to distinguish one call from the next. The chart turns many local edges into one continuous interface.

That continuity is powerful because Mediterranean coasts were anything but politically continuous. The VMN authority table uses twelve status terms across its published port phases. Some rows describe the metropole or a capital; others a subject port, colony, quarter, fondaco, tributary, feudatory or foreign port. Those words are not ornamental labels. They distinguish direct rule from privilege, possession from access, sovereignty from commercial presence. To flatten them all into “Venetian colonies” would be to replace a negotiated maritime state with a single colour wash.

The portolan performs a different flattening. It does not tell us that every harbour is politically the same. It simply has no need to say how they differ. Its users need the coast to continue past changes in ruler. A Genoese port can be essential to a Venetian itinerary. A Byzantine capital can contain a Venetian quarter. A Mamluk harbour can terminate a state convoy. The route depends on difference while the image temporarily suspends it.

This is why the chart’s silences cannot be sorted into a simple list of errors. Omitting the inland tax district is appropriate to a marine chart. Omitting the jurisdiction of the quay may be equally appropriate to its immediate navigational purpose. Yet when a later viewer treats the chart as a complete picture of trade, the appropriate omission becomes a historical distortion. We see connected ports and forget the customs officers, interpreters, consuls, warehouse keepers, pilots, creditors, guards, sailors and coerced workers who made connection repeatable.

The coast looks ready because countless people kept making it ready.

Carta Pisana · c. 1290

Anonymous portolan · c. 1320–50

‹ ›

Pietro Vesconte’s signed chart of 1311 offers a useful resistance to the political map. It gives the eastern Mediterranean as a set of coasts that can be followed and crossed. His western Mediterranean atlas material extends the same logic across a different field. Read together, the leaves suggest that a maritime network can be spatially coherent without forming one continuous territory.

Venice’s power was often like this. It did hold land, islands, fortresses and cities, sometimes for centuries. It also operated through arrangements that do not fit a modern national silhouette. Monique O’Connell’s work on Venice’s maritime state emphasises power and negotiation rather than a frictionless command radiating from the lagoon. Frederic Lane’s maritime republic was an organised political economy of ships, institutions and interests, not merely an unusually aquatic country.

The difference matters when we draw the network. A simple line from Venice to Alexandria can imply one of two stories. In the first, geography supplies a corridor and ships traverse it. In the second, the corridor exists only when a sequence of relationships can be renewed: access to a harbour, protection at a vulnerable turn, storage and brokerage at a terminus, credit in one city, information from another. The geometric line is the residue of those arrangements, not their cause.

The published VMN routes are deliberately schematic for that reason. They join authored waypoints in a stable order and avoid land at the resolution of the pinned coastline. They are not recovered tracks. They do not claim a captain passed through each displayed coordinate or that winds imposed the same curve in every season. A route line is a reading aid: a statement that these named calls formed an intelligible sequence in the project’s source model.

The distinction is easy to lose in an interactive map. Digital lines are crisp. They can be highlighted, filtered and animated. Precision of display can masquerade as precision of evidence. The responsible use of the layer therefore keeps its model visible. Ports carry dated authority rows. Waypoints resolve to stable port IDs. Operating windows are fields, not atmospheric reconstructions. Commodity terms come from a controlled table, not from the thickness or speed of an animated stroke.

Once those constraints are visible, the digital layer can do something the portolan cannot. Move the Atlas through time and the apparent permanence of the network breaks. A port changes authority. A quarter closes or is reconstituted. A fortress falls. A direct possession and a privileged foreign call no longer share the same political present. The network persists, contracts or reroutes through discontinuous rows.

The route comes before territory in the image. In history, route and territory keep remaking one another.

The Atlas below exposes the published ports and routes. Its paths are schematic corridors; select ports and move the time control to inspect the dated authority model.

Jacopo de’ Barbari’s immense View of Venice seems at first to offer the opposite of a portolan. Where the marine chart strips the city to a coastal name, de’ Barbari gives roofs, canals, churches, ships, yards and the improbable material density of the lagoon. The city becomes an object that can be inspected. Its form floats whole before us.

Yet the view and the portolan share a habit of selection. De’ Barbari makes Venice appear self-contained at the moment it was least understandable in isolation. The city’s stones depended on distant timber and labour; its markets on cargoes arriving through layered privileges; its Arsenal on organised supplies, skills and state direction. The lagoon concentrates a network whose spatial extent the print cannot hold.

Andrea Bianco’s 1436 atlas provides another Venetian answer. Its portolan leaves, tables and cosmographic material bind several modes of knowing into one manuscript. The navigator’s coast belongs beside calculations and a world-picture. The atlas does not simply zoom out from the lagoon. It collects practices by which a trading city could orient itself at several scales.

To call Venice the centre of this system is correct only if “centre” is not confused with “source”. The city concentrated decisions, finance, shipbuilding and administration, but the knowledge that reached it came through pilots, merchants, envoys, captives, clerks and travellers. Fra Mauro’s later world map makes this dependence explicit in its abundance. Reports gathered in Venice become inscriptions about places far beyond Venetian possession. The world is collected before it is controlled.

The ledger offers a useful analogy. Double-entry does not make every transaction Venetian; it makes dispersed obligations commensurable within a Venetian account. Likewise the chart does not make every port Venetian. It makes heterogeneous ports readable within a route. The achievement lies in relation, but so does the danger. Once entries share a page, they can look as though they belong to one owner.

The authority vocabulary is a defence against that illusion. Venice is recorded as metropole and capital in overlapping roles because a single status would suppress an important distinction. Candia can be capital within Crete without becoming a second metropole. A Duchy port can be feudatory rather than a directly governed subject. Constantinople’s Venetian presence can be a quarter rather than a colony. Alexandria can be indispensable while remaining foreign.

These are small words with large historiographical effects. They stop the network from hardening into an empire-shaped diagram. They remind us that an invisible map of trade is also a map of permissions.

Lines alone do not make a route. Repetition does.

The state convoy-the muda-turned a possible voyage into an institution. Ships were prepared, cargoes assembled, departures coordinated and calls ordered within operating windows. Lane’s histories of Venice and Venetian shipbuilding make clear how much organisation sat beneath maritime movement: hull form, labour, capital, regulation and the seasonal problem of getting a valuable fleet out and home again.

A portolan is well suited to the spatial half of this institution. It keeps many calls present at once. It does not, by itself, reveal the calendar that separates a viable sequence from a decorative polyline.

Petrus Roselli’s 1466 Mediterranean chart and Battista Agnese’s sixteenth-century atlas belong to a mature tradition in which the nautical image could be both working knowledge and prestige object. Agnese’s finely made atlases, produced in Venice for elite patrons, complicate any clean opposition between practical chart and symbolic world-picture. A map can carry useful coastal structure and advertise the reach of its owner’s knowledge. Its luxury does not erase its network logic; it changes the social setting in which that logic is displayed.

An adaptive timeline beside these sheets makes a simple point. Routes have windows. Ports have phases. Maps have dates. These temporalities overlap without coinciding. A chart made in 1466 may preserve inherited coastal conventions. A route operating across a period may encounter a changed authority at one waypoint. A modern dataset may encode its interval with year fields even when the underlying historical source speaks in a different granularity.

The interface should not smooth those differences away. It should let the reader see that “the route” is a family of repeated, governed passages. The line stays visually stable so that change elsewhere becomes legible: calls appear or disappear, statuses alter, a selected year activates one set of rows rather than another. Stability is a graphic control, not a historical assertion.

This distinction also protects us from false environmental confidence. The project’s source hunt found modern wind, wave and current products and millennium-scale model experiments, but no licensable dataset that could serve as an evidence-grade reconstruction of medieval sailing costs. The essay therefore does not animate winds or calculate a plausible number of days from a twentieth-century reanalysis. A calendar can be shown where it is authored; a climate surface cannot be smuggled in as history because it looks persuasive.

The convoy is invisible on the portolan not because time is unimportant, but because the chart delegates time to other practices. The map waits. The institution moves.

Commodities without arrows

Trade maps often begin with arrows and end with inevitability. Pepper flows west. Timber flows south. Grain moves towards a hungry city. The thicker the arrow, the more natural the movement appears.

The VMN commodity explorer takes a narrower approach. It uses a controlled set of tags associated with authored routes. Selecting a commodity highlights matching segments; it does not estimate tonnage, price, frequency or profit. That restraint is important. A category can reveal structure without pretending to be a statistical series.

Consider salt, grain, timber, wine, spices, textiles, metals, slaves and alum as words in a maritime vocabulary. Each invokes a different geography of extraction, demand, storage, risk and regulation. Yet on the portolan none receives an arrow. The chart offers the coast on which their movement can be planned, not a thematic account of what is moving.

This absence has two effects. First, it keeps the chart reusable. A coastline and set of calls can support changing cargoes. Second, it makes the network appear content-neutral. The same line seems available to any object and any traveller. In practice, cargo changes the politics of passage. Bulk grain and compact spices demand different calculations. Timber implicates shipbuilding and forest control. Enslaved people expose violence that the neutral language of “commodity” can itself conceal.

An interactive tag must therefore be read as an index into the project’s model, not as a moral equivalence. The category describes how a record is associated; it does not settle what kind of historical relation the cargo represents. Highlighting a route is the beginning of inquiry.

This is where the network view earns its place beside the map. A force-directed diagram discards geographic distance and pulls ports together through shared route membership. Hover a waypoint and its connections become visible. Select a commodity and the diagram shows which authored routes carry that tag. The distortion is deliberate. Where the portolan privileges coastal position, the network privileges topology.

Neither is the true shape of trade. Their disagreement is useful. A distant port can be structurally close; neighbouring harbours can belong to different circuits. The reader learns to ask what each diagram preserves and what it sacrifices.

The other half of the comparison is on the Atlas, where the same ports and routes are drawn where they actually are. Reading the two against each other is the argument; neither on its own is.

Open these ports and routes in the Atlas

Fra Mauro’s world map expands the argument beyond the Mediterranean without leaving the logic of collection. The map is crowded with text and places, turned south-up, and built from forms of report that exceed direct observation. It is not a portolan enlarged to planetary scale, yet it carries the epistemic confidence of a city able to gather distant testimony.

That confidence did not require Venice to possess the world it described. Knowledge arrived through routes that crossed rival jurisdictions. A commercial republic could be provincial in territory and cosmopolitan in information. The map makes this asymmetry visible: a world assembled in one place from relations reaching far beyond it.

Battista Agnese’s atlases later join portolan maps to cosmographic diagrams and newly expanded oceanic geographies. Their makers worked within a market for objects that could orient, instruct and impress. The network had become a cultural form. To own an atlas was to possess a disciplined picture of connection even when one possessed none of the connected shores.

This is the seductive achievement of the invisible map. It turns access into shape. If a report, cargo or ship can reach Venice, the place it came from can be fitted into a Venetian field of knowledge. The field is real, but it is not innocent. Routes determine whose testimony arrives. Institutions decide what is copied. Patrons determine which world-picture is worth making durable.

Harley’s insistence that maps be read as social documents helps us avoid choosing between admiration and accusation. The Fra Mauro map can be an extraordinary intellectual artefact and a selective product of its setting. A portolan can be an effective navigational technology and a diagram whose clarity hides labour and jurisdiction. Critical reading does not diminish the object. It restores the conditions that made the object possible.

Return to the Carta Pisana after passing through the authority table, timeline and network. The image has not changed. The same names line the coast. The same rhumbs cross the sea. What changes is the question.

At first we asked where a ship might go. Now we ask what must remain in place for the line to look effortless.

A ship must be built and supplied. A crew must be recruited, paid, disciplined and fed. Information must travel ahead of cargo. Credit must bridge the interval between departure and sale. A foreign ruler must tolerate or formalise a presence. A quarter must be defended or renegotiated. A feudatory relationship must keep producing access. A subject port must be administered. A convoy must leave within a workable window. Violence must be threatened, enacted or remembered. None of these conditions is guaranteed by the shortest line between two points.

The portolan’s visual language brackets those conditions because its purpose lies elsewhere. The danger begins when we inherit the bracket as an explanation. Modern network graphics can intensify the problem: smooth animation and exact coordinates make connection look automatic. The answer is not to abandon abstraction. It is to make the abstraction answerable to other records.

Terra Chartarum’s Atlas lets one port open into dated authority phases. The standalone network lets one waypoint reveal connected routes. Commodity filters expose a controlled thematic layer without inventing quantities. Deep links keep essay and map in a shared state. Each device adds friction to the elegant line.

That friction is historical knowledge.

Trade is “invisible” here in two senses. Much of its supporting apparatus is absent from the historical chart. Much of the network itself is also intangible: privilege, credit, expectation, schedule, reputation and fear. Yet invisible does not mean unmappable. It means that no single map should be allowed to close the account.

The portolan maps a coast. The authority table maps a vocabulary of power. The route layer maps an authored sequence. The commodity explorer maps associations. The essay maps the disagreements between them.

Together they replace the fantasy of frictionless flow with a more demanding image: trade as a route continually produced by people who did not share one status, one interest or one freedom of movement. The lines remain beautiful. They are no longer innocent.