This is the script of my presentation to the Rare Books School, University of Virginia, Charlottesville, 23 July 2018. It is largely unaltered except: slide markers have been removed, images and other material from slides have been embedded as appropriate, and there are some small edits to make the work more suitable for this medium. The only citations are those I originally provided on the slides. The talk was also recorded.
Elements of this talk will appear in Chapter 3 of Cartography: The Ideal and Its History (Chicago, 2019) and future works. Other elements replicate other posts on this site.
• thanks to Michael Suarez for the invitation and generous introduction, and to Jeremy Dibbell for all the logistical help
My goal this afternoon is to reconsider the materiality of maps and its significance for understanding maps and mapping. Some of you might find my comments to be rather old hat, as they rest significantly on the arguments of historians of the book and students of culture more generally. However, I have repeatedly seen scholars who are critically minded in one arena nonetheless surrender their critical faculties to the apparently commonsense notions of maps and cartography that permeate modern culture. The problem is that modern culture features the deeply rooted idealization that there exists a singular and universal endeavor of “cartography” that produces a coherent and unambiguous category of things called “maps.” This “ideal of cartography” comprises a complex web of preconceptions that cohered over the course of the nineteenth century.
Many scholars from across the humanities and social sciences have, since 1980, challenged some of those preconceptions and have argued that maps are cultural documents made for social functions. Even so, the sociocultural critique has focused on theorizing “the map,” and critics have not actually targeted the ideal of cartography as a whole. The result is that the ideal’s preconceptions have persisted in the academic as well as the popular mind. Not least among them, the conviction that there actually exists a category of phenomena that we can unambiguously define as “the map.” Frankly, it’s all been a game of intellectual whack-a-mole. No sooner is one preconception hit on the head than another rises up to take its place. My goal today, therefore, is to analyze and dispel the particular mole of materiality. I first consider how the ideal of cartography has promoted improper conceptions of the nature of maps as things and then to use two case studies concerning the early regional mapping of New England to reconsider the matter of materiality.
As one more piece of puffery, I should note that this topic is part of a forthcoming work, Cartography: The Ideal and Its History, which is scheduled to appear early in 2019. The book outlines the flaws of the ideal’s many preconceptions and explains how the ideal cohered over the last two centuries. The epigraph to chapter one says it all: “there is no such thing as cartography, and this is a book about it.”
Let me give you an example of what I’m talking about. It was a commonplace of early modern culture that regional maps “mirrored” the earth and that readers could use them to observe distant lands without hazarding the rigors of travel. But, after 1800, commentators increasingly accepted that all maps, even maps of the entire earth, are actually produced by direct observation, especially from an elevated vantage point. Just recently, a dealer introduced me to a wonderful manifestation of this new conviction, one that was made perhaps as early as 1820. A British schoolchild drew “the planets as seen through a telescope” which included, as figure three, the earth! How can that be? The other planets are all drawn “as seen” — such as the sun, apparently suffering from the mumps, Saturn, and Jupiter, with their moons — but the earth appears as a map, complete with lines of latitude, longitude, and the ecliptic that are actually invisible. Here, and in the rest of modernity, the abstracted nature of geographical maps is casually conflated with acts of observation. Even regional and world maps ceased to be understood as creations of the human intellect and were naturalized as direct, visual records of the world. All maps became, in the common, naïve formula, “pictures of the world.”
The ideal presupposes that maps are necessarily physical, material things that are made to hold cartographic images. It promotes a conceptual divorce between the “map image” and the “map,” between content and form, between noumenon and phenomenon. As the Dictionary of Human Geography proclaimed from 1981 to 1994:
MAP IMAGE AND MAP. The map image is a structured cartographic representation of selected spatial information, which when placed onto a storage medium becomes a map. . . . (Blakemore 1981)
(And the entry term of “map image and map” was abandoned only in the dictionary’s 5th edition, 2009.) This distinction — no matter how illogical, no matter how disproven by historians of the book in their repudiation of the “ideal copy” — has several implications that persist and distort scholarly interpretations of maps and mapping.
The crux of cartography is the algorithm for preparing map content: the observation and measurement of the earth, and then the recreation of the earth on paper at a reduced scale. The content and meaning of the map image is defined almost entirely by the part of the world it reproduces. The map image refers only to the world; for the reader, it is a means of accessing that world. The map image is self-contained and bounded by set physical limits: the image ends at the edge of the map thing. From this idealized perspective, maps are decidedly not intertextual.
In this respect, the issue of maps’ materiality is relevant only to the study of map production, to how map form might affect the look of the map image and the presentation of its content. Otherwise, the ideal pays no attention to the material nature of the map; the material existence of maps is taken for granted.
“The map” thus stands as a physical, stable thing. For Bruno Latour, maps were the stereotypical “immutable mobile,” the inscribed record of observations that can be carried from the field to the center of calculation without being changed, an act that stands at the very heart of modern Western science. “The map” also proclaims a temporal fixity. Commentators routinely note that maps are out of date as soon as they are made. The moment of the map’s completion cleanly distinguishes the epoch of the making of the map from the epoch of the using of the map.
The materiality and apparent self-containedness of maps creates a physical barrier between the two sets of individuals involved in mapping: the map maker and the map reader or map user. The study of cartography and its history accordingly focuses almost entirely on the technologies and techniques by which the map image is algorithmically structured, with some attention to the physical construction of the map, and on the assessment of the quality of the results.
By contrast, sociocultural critics have directed a great deal of scholarly attention to issues of map use, or at least to the wider social situations in which maps have been used. They have also shed light on the existence of im-material maps that feature oral and gestural semiotic strategies. Yet these new perspectives have had little impact on how scholars have sought to reinterpret material maps. In particular, the map’s materiality still appears to interrupt and break up mapping practices. In 1999, for example, the prominent cultural geographer Denis Cosgrove, having reviewed the field of map history and its traditional concerns, proposed two new “sets of questions” that should
bear heavily on any history of mapping. The first is the complex accretion of cultural engagements with the world that surround and underpin the authoring of a map, that is, treating the map as a determined cultural outcome. The second is the insertion of the map, once produced, into various circuits of use, exchange and meaning: that is, the map as an element of material culture. (Cosgrove 1999, 9)
Cosgrove outlined two sets of practice: one to produce the map and the other, once the map has been made and has material existence, to govern its use. In doing so he continued to bifurcate the entire process of “mapping” at the map, “around which pivots whole systems of meaning, both prior and subsequent to its technical and mechanical production.” Graphically, his system can be represented as:
All told, a series of convictions are sustained by the ideal of cartography’s preconception of materiality and persist despite the best efforts of sociocultural critics:
• Maps are things (but who cares of what kind)
• Maps are stand-alone, self-contained documents
• Maps sharply divide mapping into two processes: map making and map using
• Maps are stable and “immutable”
• Maps possess chronological fixity, being made at a specific moment
And we might equally replace “maps” in this list with “books.” All of these convictions are in need of rejection, revision, or at the very least substantial qualification.
William Hubbard’s Map of New-England (1677)
Let us begin to do so by considering the story of the first map printed in North America, specifically William Hubbard’s Map of New-England from 1677:
This is a famous map, long held as an emblem of the transfer of European civilization and technology to the wilds of the American frontier. (In case you’re having difficulty with the image, north is to the right; Cape Cod is at the lower left.) It also indicates the kind of detailed physical analysis that the ideal of cartography has allowed. The issue is that there exists a variant map, also cut in wood, at the same size, etcetera, but with several differences from the first variant, including several erroneous toponyms:
The most obvious difference, the name of the mountains in northern New England, gave rise to their common labels: the “White Hills map” and the “Wine Hills map.” The maps had been produced as part of William Hubbard’s history of King Philip’s War, known as A Narrative of the Troubles with the Indians in the Boston imprint, and as The Present State of New England in the London imprint (Hubbard 1677a, 1677b).
The problem was that antiquarian dealers in the nineteenth century had so mixed and matched books and maps, and had moved the maps while rebinding the books, that bibliographic certainty was lost: no one knew for certain which variant map went with which variant book. A small debate developed in the later nineteenth century trying to identify which of the two variants was to be lauded as the first map printed in the colonies. Randolph Adams (1939) published a detailed bibliographical analysis of as many impressions of the two books in original bindings as he could find; despite the small sample size, he concluded that the White Hills map was part of the Boston imprint. But his careful, empirical analysis flew in the face of the common-sense understanding that correct maps must replace incorrect ones. (This conviction stems from the ideal’s preconception of cartography’s innately progressive nature.) Richard Holman (1960) constructed — without evidence — a contorted sequence of events to explain how John Foster had first cut the block of the Wine Hills map to be shipped to London, with all its errors, and then the corrected White Hills map for printing in Boston. Eventually, David Woodward’s (1967) solution relied on the fact that the lettering in the title blocks of both variants was printed from foundry type set into the wood block. By comparing the type in each map to that in each book, Woodward verified Adams’s findings; he further explained the simple mechanics of how the Wine Hills map’s errors had been created when the London woodcutter had copied the White Hills map.
There is, however, much more to the materiality of the White Hills and Wine Hills maps than how they were made. In the very few remaining copies of the Boston edition of the book in original binding, the White Hills map was tipped into the middle of the volume; note the number of pages preceding the map. Here, at the end of Hubbard’s main narrative, the map introduced an unnumbered, seven-page section, with the heading:
A Table shewing the Towns and places which are inhabited by the English in New-England: those that are marked with figures, as well as expressed by their names, are such as were assaulted by the Indians, during the late awfull revolutions of providence.
Each paragraph in this table began with a number and the name of a place and then briefly listed the events that had occurred at that location, mostly Indian attacks on English settlements; most paragraphs provided further cross-references to pages elsewhere in the volume where more in-depth accounts of the events could be found. The paragraph numbers referenced places on the map that preceded the table; many of the towns indicated on the map bore only a number.
On this detail, “55” is where I work, in what is today Portland, Maine, which is paired with this statement:
55. Falmouth, on the hither side of Casco Bay, where August 11. about thirty-four persons were killed and taken by the Indians p. 32.33.34.
This wealth of empirical data and its distribution on this map were essential to Hubbard’s overall demonstration, contra Increase Mather, that the war had not ended because Mather had reestablished the Puritans’ salvific Covenant with God, but because God had miraculously interceded for the English at a few key moments (Edney and Cimburek 2004). In this context, wrapped up within and integral to the book, the map serves as an analytical device more than it does a locational one.
Indeed, Hubbard plainly wrote for readers who already knew the region’s geography, except for the more northerly parts. This is clear from the introduction to his secondary narrative, on the northern conflict, in which he provided a verbal, written map of the coastal settlements from Pemaquid south to the Piscataqua. He did so because the region was “less well frequented, and so more unknown” than the rest of the region. In other words, Hubbard’s readers in New England did not need the map to understand the events of the war; they needed it to understand Hubbard’s religious argument about the nature of the war.
However, when London printers got ahold of an impression of Hubbard’s book and reprinted it, they not only copied the map, to create the Wine Hills map, they also moved it. The Wine Hills map was inserted directly after the prefatory material and opposite the first page of the main narrative. Such placement was common for imperially minded English books about distant places. For example, when Increase Mather’s son, Cotton, sent his history of Puritan settlement in New England to London to be published, the printers inserted a map by the London mapsellers Morden, Berry, and Lea at the front of the first volume:
This map has since become known as the “Mather map,” even though Mather had no hand in either its creation or its placement in the volume. The Wine Hills map in situ — removed from the ostensible context of the table, and being read by readers in old England who had little access to other detailed maps of New England — emphasized the geographical stage on which the historical action took place, downplaying its analytical function and religious significance.
We can glean a couple of insights from this case study. The White Hills map, in the Boston book, began an indexical chain leading from the map to the table to the text: where does the map text end, and the verbal text start? This map is not a self-contained thing; its meaning is not dependent solely on the manner in which it depicts the world; it is not semiotically closed. Rather it is completely integrated into the larger whole, a demonstration of intertextuality. The Wine Hills map might have lost that overt indexicality, but it too would have been read as part of the London edition. More generally, regional maps are always fully integrated into arrays of other written and graphic texts. Regional maps are not read in isolation. Hubbard only made explicit what is implicit in other geographical writing, that readers are expected to move from map to narrative and back again, so that maps blur semiotically with the written word. The same principles apply to all kinds of maps: they are all semiotically open. Where do they end?
The form and content of maps are not distinct — save for the limited effects of design on content — as the ideal would have it, but they are intertwined. The physical placement of the maps in their parent volumes clearly shaped their potential for interpretation by their readers. Can we say that the two printed maps are the same? Their semantic potential is so different between the two books that I am no longer comfortable referring to the Wine Hills map as “Hubbard’s map.” And, of course, their semantic potential is completely obscured when we cut either map out of its parent volumes, whether by physical or digital means, and isolate it as a strictly spatial images.
Variant Discourses in the Early Mapping of New England
The maps of New England in Hubbard’s history suggests that maps’ materiality varies between spatial discourses, so that analysis of that materiality can help modern historians situate maps within the proper discourses within which they were produced and consumed. I’d like to explore this issue through a review of the different forms of geographical maps of New England in the colonial era and in the early Republic.
Under the ideal, cartography’s history appears as little more than the progressive increase in the quality and quantity of geographical information held in maps. This progress is implicit in historical accounts, but is explicit in the many enumerative cartobibliographies that organize maps of a given region in chronological sequence. Those maps whose content appears especially accurate or advanced for their time are lauded, and are inducted into the canon, especially those “mother” or “type” maps that set the pattern for a generation of subsequent maps. For example, John Green’s map of New England, first appeared in 1755 and was reproduced at size and in reduced formats for over fifty years:
By contrast, this cloth map will never appear in a traditional map history. It is just too simple and small:
Here it is at about the same scale as Green’s map. It simply cannot possess the detailed content that would have made it significant to historians.
More important, it is insufficiently old. As one proceeds through time, the number of maps made of an area tends to increase steadily until they can no longer all be easily accounted for. At some point in the chronology, the enumeration of maps of a region must end and attention be transferred to maps of component regions, at least while they too occur in manageable quantities. The result is a repeated fracturing of the “spatial frame” of “the map” of the region, which historians justify by arguing that it is necessary to drop entire classes of maps from consideration when they could no longer hope to show new information. Traditional map history eliminates from consideration maps that do not lie on the presumed trajectory of progress, even if those maps demand attention from a sociocultural perspective.
In the case of colonial New England, map histories and cartobibliographies have presented a “funnel-shaped” narrative of spatial constriction.
To begin with, the maps are small. Over time, the maps are supposed to have grown physically ever larger in order to hold ever more data, with single-sheet maps giving way to multi-sheet maps, or even atlases. Eventually the regional maps get just too big and become unwieldy and are replaced by separate maps of each sub-region: first individual colonies or states, then counties, and eventually the maps of the U.S. Geological Survey. Once a spatial frame is no longer relevant to the acquisition of information, it drops off the historical radar.
Moreover, such strictly chronological accounts end up yoking maps together that otherwise have little or no connection in terms of the contexts within which they were produced and consumed. The cartobibliographies pertaining to New England create a single narrative that begins in the small-scale regional maps of Renaissance Europe, and extends, through their main lists and appendices, to the large-scale topographical quadrangles and fire-insurance urban plans of the industrial age. But I defy anyone to explain how maps by, say, Giacomo Gastaldi in Venice in the sixteenth century and the U. S. Geological Survey in the nineteenth are in any way comparable!
Yoking all these different maps together in a single narrative of progressive growth of geographical content is historically inept. A first step to restoring eptitude is to consider the material nature of the maps. Doing so reveals three distinct sets of regional maps, none of which disappear but continue into the nineteenth century.
The first set comprises relatively small maps, generally about the same size as, or smaller than, modern legal-sized paper. All were published within books that either promoted or celebrated the success of the English, and more especially the Puritans, in settling the region. This particular map was published at the front of Daniel Neal’s two-volume history, published in London, on the occasion of the centenary of the Plymouth settlement.
The second set also contains single-sheet maps, but larger and on heavier paper. This map is twice as high and twice as wide as the map in Neal’s book. Rather than the multiple folds of the maps bound into books, they possess one, vertical fold along a central axis. These maps were folio atlas sheets, produced and consumed as part of comprehensive statements of geographical knowledge. These maps were not of New England per se, but of New England situated within the larger extent of English colonies along the Eastern Seaboard of North America.
The third set is of large wall maps of the region. These did not grow ineluctably out of the accumulation of information, but from contingent circumstance: a Boston-based doctor, William Douglass, wished to make a statement about the political and economic independence of the region. Douglass’s regional map is large as befits a work envisioned as a wall hanging, much larger than the maps in books or the maps in atlases. Douglass’s map did not sell and is now exceedingly rare, although for political reasons its derivatives did, beginning with John Green’s Map of the most Inhabited Part of New England in late 1755. These large maps were generally intended for display and must be analyzed accordingly, as Martin Brückner has begun to do in his wonderful book, The Social Life of Maps.
Rather than the linear narrative of increasing quality and quantity of geographical data, for early New England we need to distinguish three separate narratives, each continuing into the nineteenth century and beyond. The material nature of the maps permits us to discern not just three genres of geographical maps, all part of the same overall mode of geographical mapping, but three precise, distinct geographical discourses about New England, each with its own processes of production, circulation, and consumption. They might all be maps of the same region, constructed in the same way and likely from similar source materials, but each set featured different systems of intertextuality. The maps-in-books were integral to debates over New England as a coherent and distinctive region. The maps-in-atlases were no less connotative, but situated New England within a greater ambit of geographical knowledge integral to early globalization. As wall maps, the larger works possessed elements of spectacle intended to overawe the reader with their arguments about the economic and political nature of the region rather than engaging the consumer in reasoned debate.
Do this kind of comparative material analysis enough times, and one can discern not only such precise discourses, but also much cruder distinctions between markedly different modes of mapping. Between say, regional mapping and property mapping, or the mapping of urban places. Such maps conceptualize the world in different ways than regional maps, some circulate within the marketplace, others only within private and administrative circles. And maps are displaced, from one discourse into another. From this perspective, the unitary history of cartography fragments into a series of discursive traces that overlap in all sorts of fascinatingly contingent ways that desperately need to be investigated.
Implications for a Processual Approach to Mapping
Maps have never just gone out into the world, to circulate freely as a great, disaggregated, virtual library of images that people can be accessed at will. Not even the published maps on which I have focused today. Their material nature provides crucial information about how maps circulated, binding certain producers and certain consumers together, and so establishes their discursive context and the potential for their interpretation. Circulation — and materiality — is the core of a new approach to map studies. While the ideal of cartography has focused scholars’ attention on the production of map images, and the sociocultural critique of maps has redirected attention to consumption, and especially to the interpretation of map meaning, the analysis of circulation — and materiality — is the heart of a processual approach.
Yet maps are not interruptive of the processes of production and consumption. They are rather part and parcel of the dynamic mapping processes of production, circulation, and consumption. Circulation is key, as it’s the process by which production and consumption are integrated. We can reconstruct some of general parameters of specific circuits through which certain sets of maps circulate through the analysis of their material qualities. I have changed the color of the arrows to emphasize that the flow is not of information or content, but of the maps themselves. Participants in the discourse can take on many roles as producers and consumers
Material maps might be physically stable and fixed, but their meanings are not. They are not out-of-date as soon as they are made. Map consumers construe not only what a map’s symbols denote, but also what an entire map connotes. People are always undertaking mappy acts: making maps, circulating them, using them. As maps continue to circulate within their discourses, or cross between discourses, and are found to still be meaningful, they remain valid and up to date. Even the storage and destruction of maps are dynamic, requiring decisions to be made and actions to be taken; archives and libraries are in fact not just places of storage but sites of further knowledge production.
This simple diagram is emblematic of an entirely new approach to the study of mapping, an approach that decenters not just the nonsensical category of “the map” but maps generally. Maps are epiphenomena of mapping processes. At the same time, they are the primary things through which we can study those processes. In this respect, the study of the variable material nature of maps — including their immateriality — is a crucial component of a processual approach to mapping and map history.
As the self-containedness of the individual map is debunked, as its edges metaphorically fray, so too does the generic category of the normative map. I, for one, have a hard time treating these two works as part of the same category of phenomena! And with the idea that there can be an unambiguous category of “maps” being thrown out the window, with it goes any justification for the concept of the singular, transcultural endeavor of cartography.
There is no single endeavor of “cartography” that can be traced across human cultures across the centuries. There is no unambiguous category of things called “maps.” What there have been, and what there are, are innumerable spatial discourses in which people conceptualize the world in different ways for different purposes; those discourses entail a diverse array of mapping practices, producing maps with an array of material, and immaterial, semiotic strategies. There is, however, sufficient consistency that the analysis of maps’ materiality enable us to meaningfully arrange mapping discourses as a series of intersecting mapping modes.
Adams, Randolph G. 1939. “William Hubbard’s ‘Narrative,’ 1677: A Bibliographical Study.” Papers of the Bibliographical Society of America 33: 25–39.
Blakemore, Michael. 1981. “Map Image and Map.” In The Dictionary of Human Geography, ed. R. J. Johnston et al., 199. 1st ed. Oxford: Blackwell Reference. Repeated with little change in 2nd and 3rd editions of 1986 and 1994.
Cosgrove, Denis. 1999. “Introduction.” In Mappings, ed. Cosgrove, 1–23. London: Reaktion Books.
Edney, Matthew H., and Susan Cimburek. 2004. “Telling the Traumatic Truth: William Hubbard’s Narrative of King Philip’s War and his ‘Map of New-England’.” William & Mary Quarterly 3s 61, no. 2: 317–48.
Holman, Richard B. 1960. “John Foster’s Woodcut Map of New England.” Printing and Graphic Arts 8: 53–93.
Hubbard, William. 1677a. A Narrative of the Troubles with the Indians in New-England, from the first planting thereof in the year 1607, to this present year 1677. But chiefly of the late Troubles in the two last years, 1675 and 1676. To which is added a Discourse about the Warre with the Pequods In the year 1637. Boston: printed by John Foster.
Hubbard, William. 1677b. The Present State of New-England. Being a Narrative Of the Troubles with the Indians in New-England, from the first planting thereof in the year 1607, to this present year 1677: But chiefly of the late Troubles in the two last years 1675, and 1676. To which is added a Discourse about the War with the Pequods in the year 1637. London: Thomas Parkhurst.
Woodward, David. 1967. “The Foster Woodcut Map Controversy: A Further Examination of the Evidence.” Imago Mundi 21: 50–61.