Category Archives: choropleth maps

Variations on Our Fragile Union: Small Stores across 40,000 Zip Codes

Any maps create quite powerful tools to knit regional divides into coherent forms.  The order of maps serve both to fabricate harmony of a whole from its parts and for showing its divides, in ways data visualizations make increasingly clear.  If the origins of “chorographies” were illustrations of communities, and aestheticizations of their harmony, we’ve become increasingly interested in and attuned to visualize divides, as much as visualize unity, to explain the symptoms of divisions where they might be most evident, and use the format of the map to demonstrate compelling evidence of the fractures that divide, and explore how ideology ever came to “cluster so predictably,” in with geography,” as Steven Pinker nicely put it, and if different conceptions of human nature exist in different parts of the country. We’ve all been all too familiar with the recent remapping of these divisions, increasingly evident  in our political discourse, and affirmed in different ways by posing a set of choropleth maps as points of departure in order to chart, gloss and explain the extent of our divides by the range of data available on local differences across our fragile union.

As we approach Christmas and the season of gift-buying upon us, it makes sense, as Philip Bump, with some help from folks at Esri and Stamen, suggests, to consider the stores where we realize shopping habits and alternatives as criteria that increasingly divide us–and consider the consequences such divisions hold.  Indeed, the divided commercial topography in many regions of America reveal–even more than real estate–a deeply drawn set of cultural divides rarely considered in attempts to understand our fractured political landscape:  so dense are the clustering of large stores in certain areas that they obscure the legibility of place names.  In ways that reflect something like a Google Maps query “shopping at private business near [ZIP code]”, the distribution of data suggests how the landscape of shopping shifts in regions of the United States that provide a deeply qualitative record of our society.  But rather than generate a list of sites, or map all stores, selective parsing of large scale vs. independent shops provides a neat filter to measure the commercial contours of the landscape in ways that won’t so much correlate as reflect a composite image of the lower forty-eight. What’s the new political topography of the United States?  Well, perhaps it’s less red states vs. blue states, here shown in two views of the divisions between Democrats (blue) and Republicans (red)–by data here parsed at county level and by intensity of affinities–map a deep set of social divides that have strikingly played out in weirdly clear geographical blocks in presidential elections.

As Steven Pinker put it, “Regardless of who wins the presidential election, we already know now how most of the electoral map will be colored, which will be close to the way it has been colored for decades.”  Why, indeed, are some red states so red, and other states so blue?  Is this a sign of a divide between rural and urban inhabitants and political preferences–or is it deeper, and, if so, how did we become so divided in our imaginary of what the nation should be?  Are these maps creating differences, or do they reveal some of the deeply-set social divisions codified by Mason and Dixon as a latitudinal divide, but whose division is now reflected ban urban/rural divide? We all know that the last few elections have been stunningly fragmenting–

votes-  red v blue, by county and interest level

In a coy reference to collective mapping, Aaron Black of the Washington Post used Gallup’s annual rankings of which states are the most conservative and which are the most liberal to create a heat map of populations across the nation, employing the popular cartographical device so often adapted in infographics to chart virtual a spectrum of political opinion, which, while less intended to emphasize the fixity of divides within current public opinion, showed that outside of Alabama and Wyoming, the number of self-declared “conservatives” in the country does not follow such a close divide:  yet it is hard to find a less changing term than “conservative” and “liberal,” and these don’t map onto the breaks in voting that have recurred across four presidential elections.

%22Open Heat Map%22Washington Post

But the pressing question of where these divides actually lie have led folks to return to maps with considerable zeal–both as powerful heuristic vehicles of statistical measurement and projection, and to find possible answers for these real divides.  The prevalence of infographics and talk radio have conspired to make this a real issue of political debate for the The American Conservative, where Patrick J. Buchanan placed the question of secession on the front burner this year, corralling red state secessionists and blue-state secessionists into the same category as folks who acknowledge we either face increasing ungovernability due to geographic expanse or having little in common:  the issue of what we have in common, not much of a question for the very notion of a nation or state was in the past understood to balance multiple interests, may be less the question than the “spirit of intransigence” that we face in most political debates.   Does this have to do with the vociferous defense of issues that won’t be debated, or the fact that we have been all too often poising debates in quasi-imperial essentializations of what is un-American?  Or is it due to “lifestyle”?

Can we remap these divisions along lines of population density, as Robert Vanderbai of  Princeton University has argued, whose compellingly recast the electoral map to reveal how two alternate styles of sociability came to inhabit one land.  By throwing the stark red/blue political topography into 3D relief, Vanderbai invites us to look at that divide, not fractured by different data, but mapped onto density so that we might better understand our electoral maps:

Blue v. Red Topo Raised

The divisions in the electorate have also, of course, encouraged the weirdly popular but loosely credibly parsing of local attitudes across the nation.

Despite the hugely informative value of this density distribution, the question of how the distribution of the landscape we have created might be as well illustrated by the commercial topography we have created, and the space that they create.   For the interesting map that charts a distribution between the majority dominance of large box-stores or chain-stores over smaller privately-owned stores suggests a unique landscape that we have become all too used to inhabit, and a measure of density all its own–but not necessarily in tandem with Vanderbai’s glorious and striking distribution of a two-color bar-chart.  Although Vanderbai’s chart makes the compelling case that our distribution of electoral votes no longer in fact reflects the population distribution in the country–and indeed the possibility of the winner of a popular vote losing the number of votes in the electoral college–the map is less than informative of the landscape we have created in the United States–a landscape that is still being repeatedly surveyed.  To be sure this reflects or is a reflection of density, but also offers an alternate index of social topography.

Let’s peak at the regions that Bump has created, in far more fuzzy fashion, by agglomerating the number of stores across its 40,000 zip codes, and taking a measure of those regions where large retailers show a clear dominance to the extent that they have pushed out all but one privately owned retailer, noted in reds of varied intensity, versus those blue dots where privately owned retailers constitute the majority, and the corridors that this creates, despite its lack of clear contours.  While the ZIP-code breakdown is reflective of the new concentration of wealth on the eastern corridor, the relative decline in incomes provides a clearer index of where such box stores (predominantly discount retailers, who appeal to the pocketbooks of their customers) will tend to concentrate.

Fuzzy USA

Income by County:2012

The contours of these divides provide a basis for re-examining the arguments implicit in most recent choropleths, and thinking about the nation we are poised to become.  Although not exactly mirroring the levels of declining income, one can detect in the areas of the spread of the big red superstores a reflection of the yellow states with median incomes below $22,126 a year, showing a selective density of big bog scores pushing in toward Denver, and congregating around Michigan, the Great Lakes, Iowa and near Arkansas.

Red Crawling INland to Denver

Starting small, for example, take the data that Natural Earth and OSM mappers help reveal about the new topography of shopping in the Midwest, that both reflects density but even more the sorts of incomes that can support private businesses:  as one moves outside of urban areas, the proportion of zip codes composed almost or entirely of exclusively large retail stores rises in clusters, often clustering (predictably) about pockets in precisely those cities also able to support small businesses:

Density?

The blurring of the map, in a far fuzzier version at larger scale, suggests a veritable field of red that is almost aligned in belts, and creates a strong contrast to many parts of the country lying further west . . . where they seem to choke off privately owned stores.

Fuzzy Mid-West

Returning to view the southern reaches of the country a greater scale, the discreet stores in much of South appear as oases of small businesses in relatively open land–presumably where small businesses don’t crest above a level of 50%– in Nashville, Atlanta, and Memphis:

Southern US

Looking at the odd distribution of the Eastern seaboard, a region where large stores appear to dominate the landscape of all but the coastal corridor, and gain a distinct density as one moves inland and approaches a different economy.

Fuzzy NE

A more fine-grained view of the same region–focussing on New York to Baltimore–reveals a more complicated topography, but echoes the general division between small pockets of private stores and box stores that choke out the competition, being the chain stores that they are, which, with relatively cheaper real estate at malls that can be bought up and leased out by developers, creates a new division between our populations that may be deeper than–if it is created by–a gap in income or disposable wealth.  Interestingly, an abundance of privately held, smaller stores is often concentrated around smaller towns, as well as taking on density at large cities.

New York to Philadelphia

What sort of social topography does this create?

The distribution of stores has not been indexed for growth rates, but might be measured against recent census findings about declining incomes.  The depressing image of the appropriately colored “brown” marking declining median household incomes in a period of just five years nicely captures the tone of depression of much of the country, and the difficulty of political classes to capture much backing outside the venting of folks whose incomes have declined.  Despite the aversion of a serious depression after the banking crisis, we’ve been hardly hit.

Median Income Changes, 2007-12

But the choking off of private businesses might be a sad consequence of these declining incomes, and the road trip searching for big bargains at Dollar Discounts that might be taking up the weekend breaks of much of the country.  While demanding further investigation, the number of chain stores that have emerged across in the country surely reflect the declining incomes in the nation, as much as changing costs of real estate.  And these m

The proliferating choropleths that have emerged to explain our social divisions, something of a tour of data visualizations that have duplicated like rabbits on the web will reveal, have in fact tried to uncover divisions in moods as dividing the country–although this map, with the authority of a DSM, seems to assign fixed moods to the boundaries between states that we have come to internalized, so nurtured are we in reading data visualizations as records of objective truth.

On one extreme, the heuristic value of those boundaries between states are fetishized in a spectacular fashion in America’s Mood Map, which uses statistics originally published in the Journal of Personality and Social Psychology to detect underlying differences of constitution–the personal, rather than the political–that creates an odd visualization whose  very terms (“relaxed and creative”; “friendly and creative”; “uninhibited”) make it sound like we’d all really get along quite well, aside from difficulties of understanding tone, and downplays any political division by recasting the map in terms of something like affect.  By ostensibly mapping divides of personality types and tabling the political, this map even invites you to find the state that best matches your mood, if one is to believe self-reported data.  (The division was parsed, with a straight face, as reflecting weather–hypnotized by the ubiquity of the weather-map and seeking to invest something like objectivity in the odd color-scheme below as something that revealed actual differences to invest the odd graphic with similar credibility–for more on the notion of mapping how “different regions have different personalities” in a data visualization, look here.)

America's Mood Map

 

Regional essentialization has proceeded, in an attempt to map the country as made of different nations.  In the drive to map national divides, Frank Jacobs parsed nine nations of North America on largely economic criteria, rather than personality, distinguishing regions of similar economic production, and casting each as a separate GDP, based on concepts of the uniqueness of our nation from WaPo journalist Joel Garreau, and again granting them stark objectivity via a choroplethic map that similarly served to instantiate diverse regions of the continent:

 

GDP map Is this a proposal for introducing new tariff lines?  If so, how much such a distinction makes sense–what about that “Empty Quarter”?  what on earth do they do, and how can they exist?–can be tabled.  And the future of the “foundry,” if it rests on metallurgy, is certainly on the rocks.  And what does Quebec produce, aside from fleurs-de-lis? Colin Woodard sought to describe national divides into some eleven groups, arguing that we’ve “we’ve never been a nation-state in European sense,” for all who imagined the possibility of American political unity, and wrote a slim book that recast the Progressive notions of a ‘melting pot’ by explaining that each region reflected its own distinct stock, and illuminate differences to which our democracy, supposedly, had unfortunately rendered us blind–but whose divisions Woodard appears to suggest might explain all American history, including the Civil War, up to the spread of the Tea Party.  Woodard’s pedantic mapping of distinct regional variations combines racial essentialism and a “revenge of geography” to point up our innate fragmentation: ELeven Nations

The distinctions were starkly mapped in alternative iterations of the “American Nations” were presented as a key to better understand ourselves, as if to hold a mirror to the land:

 

American Nations Sean Wilkinson Design

 

Sean Wilkinson Design Woodard’s work extrapolated from the historian David Hackett Fischer’s argument about the differences in the settlement of different regions–particularly the northeast and south–by different Anglo cultures, and the effects of Anglo farmers who settled the northeast from the Scots-Irish herders who settled the Deep South, and from the work of many anthropologists and the psychologist Richard Nisbett on the culture of honor in the South.  But despite some  neatness in registering cultural differences in the map, it doesn’t take huge skill to reveal the culture of separateness in the Deep South, and a floating barrier of “Appalachia”–the ‘light South’–around it;  yet problems arise when one moves from the map to individual experiences. Despite the evocative names with which other regions are endowed, from Tidewater to the New Netherlands, their relative stability or homogeneity as entities comes to break down, as does the notion that their boundaries are so firm.  The odd coherence that he supposes in such regions as “New France”–would Québécois see anything similarly to folks in Louisiana or even New Orleans, let alone accept them as affines?–and other regions–how isolated and cut apart is this “Left Coast” today?  What could ever be mapped as weirdly as Yankeedom, and is it still alive?–seem animated by an ethnic essentialism belied by the shifting map of ethnic divides and affiliations in the United States.

And it is not clear that these areas will retain meaning with and over time–although the deep danger of which the map is something of a symptom is a pronounced lack of national belonging across the land!  Woodard’s sense that American history “all goes back to who settled those regions and where they came from” flattens history, and would not be readily recognized by any local inhabitants. No doubt these can be traced to the deep and unavoidable social divides and fissures that exist in our country, and to the different parts that different regions have played in the culture wars.  Perhaps it might make sense to adopt an approaching demographic forecasting, less entrenched in the categories used today, and the apparent generation of megaregions.

 

 

2050_Map_Megaregions2008_150-thumb-615x409-106683

 

There are, of course, true socioeconomic divides that fragment the country in all too real ways.  The frighteningly influential role of real estate in shaping the country is sadly evident when we map residential space by income, and illuminate truly early modern concentrations of wealth.  Using the American Community Survey to map a choropleth by variables of income, real estate, and educational levels attained, we can perform something like an archeology of the national map, digging deeper to examine the real divides among its inhabitants.  Dividing regions of the country by median level of income, clear pockets of navy blue surprise the eye and remind us about the unavoidability of underlying social divides–that, I’ve elsewhere argued, mirror local reception of the Affordable Care Act.

 

income

 

Or, again with an eye to the opposition to the ACA, long after the Great Society, we can divide our choros by clearly drawn concentrations of poverty as much as wealth, using a distribution of  local populations living below the poverty line in each of the nation’s counties that reveals a similar concentration of poverty in the deep south and southwest:

 

poverty

 

Not to mention the map–which somewhat scarily illustrates parallels–of the percentages of adults over 25 years of age who did not complete High School, in a sort of national map of educational attainment as they are reported in US Census figures:

 

Over 25 w:o High School Diplomas?

 

These stunning choropleths, courtesy Calvin Metcalf, Kyle Box and Laura Evans, reveal a Southwest and deep South we knew, but weren’t ready to acknowledge.  They tell an interesting story as to why Virginia and Florida belong less to the Old South–or to what it has become.   To look more closely at a comparison of other local trends in the Community Survey, look here. Yet we might do well to look elsewhere to map national divides, and new databases to grasp our emergent geographic divisions and continental divides.  While real estate markets have much to do with the divides, we seem to have individuated the intersection between density and real estate across state-lines, by taking the presence or absence of small businesses in our communities as a sort of tipping point.  For the benchmark of zips with one or more privately owned businesses reveal a major fracturing of the local landscapes in the country, oddly resonant with our divided electorate.  The huge regional discrepancies of coasts and interior needs to be read with density, in ways that few openly consider, as a means to map the habitation of the landscape, shifting from questions of bodies, ethnic composition, or other criteria, toward the nature of interaction between self and place that shops afford, and even the sorts of relations to objects that are synthesized in the contractual relations of the marketplace, not to mention the stimulation and relative criteria of judgment that a diverse marketplace affords customers. The below map of the nation is limited to those spots with no or one small businesses within the ZIP in question (the red dots, which often cluster and overlap, and are more intensely red to reflect the intensity of that concentration) and zip codes which are distinguished by a majority of small businesses (blue dots).  While this omits large regions of the country–where there are an intermediate percentages of small businesses, and they don’t register–those are of course the very regions where large stores are equivalent or exceed small businesses, making them likely sites of the inevitable large box stores.  And the selective optic of the map is precisely its virtue as a portrait of contemporary agora, or our evolving landscape of exchange.

 

small shops:large shops

 

In ways of considerable sociological significance, the map reflects the landscape we’ve made for our country, or we make for our selves, by mapping some of the sites where we define ourselves to what is left of the public sphere.  The map may indeed suggest some interesting sociological detail about where the country is heading, with interesting correlations to level of educational completion below which it lies.  There is something that goes beyond or beside a decay of discrimination, but leads to a shift in one’s relation to the market and how one is a participant or spectator in a collective activity, in the geography of box stores that can be mapped across America.   The dominance in a mental universe of Home Depot, Walgreens, Wal Mart, Dollar General, Target, Lucky, Thrifty, Family Dollar, Dollar Tree and even Sports Authority exposes a topography of disorientation and dislocation, rooted in the offshore production of goods and plastics, undermining face-to-face encounters in 24/7 Emporia that offer the best deals may well conceal endemic conditions of placelessness and timelessness:  in a country where 80% of the stores nearest to anyone are large retail chains, one has to interrogate the ramifications of what sense of social space we retain:   as we become consumers in another Holiday Season, bracing ourselves for sales, what sort of social change lies in no longer having a corner store? Where’s the analysis of this shift of commercial topography when you need it?

In ways rich with sociological significance, the map tells much about how store-goers orient themselves to the publish sphere, direct consciousness to the relation of goods’ producers, and regard the social contract or social nexus as a given or negotiated part of the status quo.  For as much as it lies in our pocket-books, perhaps the deepest divide among us of all lies in the landscapes where we spend it.  Perhaps the deepest divides of all rest on not so much in levels of income or consumption across the lower 48, so much as the shops available for perusing or spending time–and the attachments to place, to space, to objects, and indeed behavior pursuant from them, which inevitably cascade from how much time inhabitants devote to patrolling the aisles of brightly lit box stores of Sam Walton–rather than looking in at local businesses to greet their owners and examine their wares.  Based on a search for the places where “small business really thrives” in the lower forty-eight among the nation’s 40,000 zip codes, data from OpenStreetMap has allowed a map of the sorts of stores in each–noting in red those zip codes hosting none or one small businesses, whereas those where small businesses constituted an outright majority enjoy varied shades of baby blue.  The resulting small-business texturing of the country provides a valuable index for looking at–or reading, if not actively interpreting–the nature of its communities, and an alternate American Community Survey of its own.

While this doesn’t exactly correspond to a political divide, it may tell us more about what sort of a nation we’ve become–and what sort of commercial contacts that we value, or personal contacts we are able to cultivate to the owners of shops, places more likely to apt local products.   The division of red vs. blue is less reflective of a political divide, but seems to offer itself as an alternative metric to those above introduced to measure social divides, and belongs in the company of other social indicators taken as prognostics of political divides in the electorate, if not of an alternate community survey.

 

small shops:large shops

 

In ways that do not map directly onto the distribution of real estate–shown in the first map from data of the American Community Survey above–as one might expect, the topography of shopping that it indicates suggests deep social divides to which Nate Silver might pay some attention:  while the coasts are brightly blue, in California we can detect the conservative central valley and interior; the absence of blue in much of the central states is shocking, and preponderance of red throughout the south–save coastal Florida–overwhelms, although the Northwest seems reliably blue.  In Minnesota and Wisconsin, the red blotches shade to pink and dissipate; Texas seems divided with deep red concentrations.  Indeed, the problematic position of Pennsylvania in the national picture is nicely illustrated by its deep red isolation, whereas privately run business dominates the Northeast Atlantic seaboard, particularly north of New York, as well as in much of New England:

 

Pennsylvania?

 

The irony is hopefully apparent, given that the defenders of Republicans are presented as the champions of free enterprise.  Maybe they are:  they are not, it seems, defensive of independent shops or businesses. The sea of red in the middle of the country, but seems a growing miasma that doesn’t necessarily respect political affiliations, but also bodes pretty badly for the survival of a culture of localism, at least in terms of the future (and profitability) of independent business:

 

Growing Miasma

 

The map, in the image it offers of the Midwest, reveals a striking clumping of regions without independent businesses–clumps not necessarily paralleling cities–the like of which is probably a major shift in our society the likes of which the world may have never seen:

 

MidWest

 

Those places with greater income may simply be ready to foot the extra bill for local boutiques, some will say, or forget the offers of Amazon’s free shipping:  but we see a deeper divide, I’d argue, shaped by the market as much as responding to lines of wealth, and conditioned by all those aisles of shopping centers illuminated by starkly lit flourescent lights, lured by the slogan “Save Money.  Live better,” and names like Dollar Tree and Family Dollar that remind shoppers to keep the bottom line first and foremost in their shopping experience, rather than cultivating stores that might provide more jobs to their neighborhoods.   (Even if their parking lots may use LED lighting, most stores have a sort of blinding or blanching flickering flourescent light so recognizable that it seems potentially addictive as a unique neurological experience.) We can see a new mapping of the landscape around cities, which, if we take only the midwest, emerge as something like islands with private stores, surrounded by areas of box stores that cluster around the nearby highways and urban belts, to take Chicago and St. Louis; smaller towns are dominated by zip codes dedicated to large chains:

 

Around Chicago

St. Louis

And in the deep interior, many zips feature the near-absence of anything but larger super-chains:

In Illinois

And we can see the fracturing of a political landscape in divided states–such as Colorado–where diverse concepts of the geography of sales, as it were, present different models of a lived landscape, almost cheek by jowl, alternating between spaces of chain stores and a telling density around Boulder, Denver, and Colorado Springs–with a few exceptions–of small businesses.  In some ways, this reflects the deep debates in the state on such issues as gun-control, birth-control, ecology and a more nuanced or complex relation to federalism.

Complicated Colorado

 

 

Where we shop is increasingly what we are not only because of our attachment to objects or commodities, but how we are oriented to the world through the market–and as we map our relation to markets in different ways,  the ongoing onslaught of this attack of the big box stores has deep repercussions for our society’s unity.  Go forth and map as you do your holiday shopping . . . To be honest, some of the maps derived from this data don’t make immediate sense to me, or at least jibe with my own experience, such as the large number of small businesses found in Las Vegas, although I suppose this reflects the definition of a “small business.”

 

Las Vegas?

But the image of the Bay Area is somewhat recognizable, and recognizably blue.

 

Bay Area

Yet a broader landscape and context emerges as one expands to adjacent ZIP codes in order to include bordering strip malls, including those inevitable and all too visible nearby strings of outlets, and sections of highway that are dotted exclusively with box stores.

 

Bay Area Large

Leave a comment

Filed under American Community Survey, choropleth maps, Mapping Box Stores and Small Businesses, OSM, Red states v. Blue States, Red States/Blue States

Data Creep

The relative onslaught of poor data visualizations so plaguing much of the news media may derive from a hope to attract new audiences as budgets shrink and bureaus decline:  by boiling down a “story” by dispensing with those bothersome words, they seek to make immediate impact on an audience by a powerful (and eye-catching) graphic.  Based on the self-reported responses to the “Big Five” personality test questionnaire that was developed in the 1970s, but recently used to aggregate responses via Facebook, which posits “five dimensions of personality” to distinguish personality types, based on the odd belief that, rather than reflecting individual character, one could detect “different regions of the U.S. have different personalities.” The self-reported rankings of attitude (curiosity, energy-level, tenseness, quarrelsomeness, forgiveness), efficiency (reliability, laziness, perseverance, efficiency), and character (shyness, moodiness, distractibility, sociability, rudeness) are values with little possible quantifiable relationship among themselves, which translate into a data-distributions of limited legibility or credibility after they’ve crept into a map.  Projected onto a map the colorful choropleth offers a “mood-ring for the nation” whose choice of hues communicates little intuitively:

state-map-personality-test

Unimaginative data overlays like this  lie somewhere between video games, a MacPaint program, and an adult coloring book approached with Prismacolor markers–more a diagram than a map, they serve to carve the nation into clear blocks as if this would clarify anything about national unity or collective networks:  such visualizations take pride in how they disrupt continuity in a search for a narrative about the national divides that are revealed in our political process, and do so with varying degrees of precision.   Their production seems to be driven creep of data into overlays atop base maps, as if to awkwardly digest the familiarity with data–and make all feel like they have access to truly “big data”–by using an image of the nation to bequeath authority even to miniscule data samplings by treating them as images able to visualize datasets:  this is an insidious format makes us thirsty for more of the same, as we seek to grasp divides and parse divisions with the apparent exactitude of a surgical scalpel.

The recently widely retweeted but fairly facetious map of “America’s Moods”, an interactive graphic mapping emotions titled “America’s Mood Map,” has circulated online with considerable popularity but is able to be blamed on Time magazine’s website.  The data visualization has just the right mixture of declarative insouciance and light-heartedness make it a meme and bane of online journalism, and a typical illustration of this dilemma.  The distribution of data that results deflect scrutiny from the very data that they’re employed to embody.  The interactive blocks of color in what seems a choropleth distribution are a bit compelling, until one asks what state-lines have to do with emotions after all, or if this just was a nifty way of converting data to visual form.

What sort of embodiment of data is going on here, one might well wonder, and question what the mosaic of colors communicate or signify.  Not to mention the map’s confusion of a question of individual psychology and gross geographic regions–especially such abstractly construed categories as the legal boundaries of forty-eight individual states’ authority in our nation’s union.

The interactive ‘map’ demonstrates the recent discovery that responses to the great American greeting, “How you doing?,” differ starkly across state lines in the lower forty-eight:  if in benign fashion, the result proclaims divisions and splintering that trump the continuity of territorial maps, and perhaps map an explanation for all the differences we already know.

 

 

America's Mood Map

 

 

Why “friendliness” is signified by red, “temperamental and uninhibited” by blue is as problematic as the lack of any continuity among these personality types, and the relative subjectivity of judgment:  it turns out that these are self-classifications, anyway, rather than determined by objective criteria–as if values like these could be objectively assessed.

The lack of material references in such ‘maps’ almost winks to viewers not to take them too seriously.  Yet the relative ease of converting statistics into overlays on base-maps in web-based formats, seems the rationale for their popularity as interactive media in on-line news publications.   Forget the actual map that orients its viewer to the lay of the land:  this is immersion in the map as interactive data environment.

The deepest difficulty of this data visualization may lie in how it confounds the empiricism of a map with pretty relative–and pretty vaguely construed–psychological categories. Although Time magazine science editor Jeffrey Kluger seems to have fun downplaying is meaning at the same time as he promotes it, “America’s Mood Map” is the most popular in the section “Science and Space” among readers of Time this past week, and a success by journalistic standards, is the interactive map of emotions across the United States, across which one can glide one’s cursor to reveal a virtual version (and modernization) of the early modern Carte de Tendre over which you can mouse about to find a place that “matches your personality”:  but rather than visualize material renderings of feelings or emotions, as that topography of amorous practices, the imaginary topography over which we mouse to find the ranks of each state’s inhabitants reveals clear divides rather than a detailed qualitative record.  Data has crept into this map’s bright mosaic of colors can’t help but engage other data-vis maps, with which its full-spectrum color schema stands in such stark contrast.

 

Moood Map of US

Although the color blocks are arranged in something like a spectrum of friendliness to temperamental, the actual values on which they are based provides something of a map of mental constitutions, as much as emotions, and one can range of neurotic to extroverted, with open-ness thrown into the mix.  The explorer of the map can find themselves, for example, in “agreeable, conscientious and open Tennessee;” we all know a few who fit the description:

 

Conscientious Tenessee

 

The ranking of each state surely increased its popularity, as the map becomes yet another tabulation of characteristics after one mouses around a bit on its surface.  California, predictably, is both relaxed and open (#2 nation wide!) and low in neurotics (#43; agreeable Utah lies at the bottom of the heap at #49), and New Yorkers are temperamental but ranked as among the most open (#3).  (Such classifications based on a sampling of 30,000 must conceal the detailed nature of the questionnaire.) Who would have thought that largely rural Wisconsin, a state with one large city, possessed the most extroverted population in the country? Or that Maine stood near the nation-wise apex of neuroticism?  New York gets pretty low marks for “agreeableness,” whatever that means (#48 in the nation), if it is also pretty high in “openness.”

There might be some problems with the data pool.  Perhaps the map’s very lack of materiality makes it difficult take seriously, even if the pleasure of using moods to divide the country seems a relief from dividing the nation by ideological divisions.  (The next step that this map seems to invite is no doubt for carto-data-crunchers or map-readers to map the moods of the nation onto those political divisions:  how better to easily explain the ideological divisions that grip our media on the eve of the Affordable Care Act and the morning after the Government Shutdown?)   Indeed, the interest in the “mood map” among Time‘s readers might been generated in part by hits from all those readers, long subjugated to an onslaught of data visualizations, who want to explore their own states in the mirror of their own states of mind or who want to try to map the now-tacit maps of national division onto the far more innocuous (and un-ideological) question of moods.  Indeed, this stepping out of the recently emerged graphic lexicon of ideological division and splintering is somehow reassuring, as, much as the article announced, maybe its mistake  this country “features the word United in its name,” since “we splinter along fault lines of income, education, religion, race, hyphenated origin, age and politics.”

Maybe it does really all boil down to constitution and emotions, all those earlier data distributions be damned.  The end-product is something of a polemic rebuttal to the authority of earlier data visualizations in the news, to be sure, of a very tongue-in-cheek sort of very, very muted irony.  The text’s injunction to find where you belong in the map–by your mood, not by where you actually are–invites you to glide your mouse over a map with the authority of a spatial distribution of the rainbow colors of a mood ring, in a pretty abstracted state of mind, so unlike the ways in which, say, a detailed topographical map registers the measurement of physical elevations by exquisitely exact orographical detail.

The survey employed was based on a sample of under 30,000 respondents, but passes itself off as a pretext for self-examination or -understanding, complete with the assurance that results won’t be reported or stored by Time is respectful of your privacy (perhaps to marketers of antidepressants?).  Whether it is able to map such stark divisions of “mood”-tendency beyond statistical error is unclear, although the almost spectroscopic division of the nation into stereotypes seems somewhat persuasive:  the center of the country, if not so large a swath as the “red-states” of Bush years, is proudly “conventional and friendly,” unlike the creative types on both coasts:  the mapmakers permit little constitutional overlap among these categories, or multiple combinations of them, so much as render one of the three criteria for each state, and allow little overlap among them; the cartographical “paratext” to the map placed above its panels invite its readers to take a short test so that one might place your personal constitution where it really belongs, and suggests that these three metrics are rigidly exclusive from one another.

 

Moood Map of US

The result is a new portrait of the dis-united states, several of which are already in widely circulation–and some even so widely internalized as ideological divides that one can’t make associations between this “map” of emotions to more familiar political and social divisions.  The data visualization may be taken as a pretty light-hearted response to our dramatically increased geographical mobility, or our obsession with data-visualization maps.  But Kluger and co-author Chris Wilson use the data of fellow-American Jason Rentfrow, obtained at Cambridge by a multinational think-tank created data by a psychological survey of their own device, and the map is presented in the rubric of the “Science” section of the magazine’s website.  The data that was used to inform the visualization, under the name of science, claims to reflect the salient divisions of what “for a country that features the word United so prominently in its name, the U.S. is a pretty fractious place,” as if it might be a more credible set of criteria to ascertain relative depths of fractiousness and their causes–despite its odd metric for measuring “emotional” divisions.

And its interactive features create at least half of the fun for its readers.  The notion of locating diversity in our moods is a lot more appealing than finding it elsewhere; the mirror of the interactive map is no doubt a partial reason for its popularity.  Indeed the invitation to guide oneself to one’s own and one’s nation’s emotions might be hard to pass up, if it suggests quite a lack of complexity in the terrain revealed by introspection, which seems, here, to be equivalent to the completion of a modular form, rather than offering a topography that might be worthy of future qualitative detail.

There is a more authoritative, and perhaps more familiar, map of which the map dissected above might be called the comic repetition.  The study of state-specific variations in happiness (one emotion–that’s a better concept already) was the result of a study based at UVM of geotagged tweets, published in the online journal PLoS ONE, whose tabulators ranked over 10,000 words on a graduated scale to score some millions of tweets across the country, irrespective of their context, to reveal significant differences in sadness and happiness across the nation, perhaps better translating what might be called a set of emotional divides:

 

Happiness Score in One Map

 

Indeed, the clear “sadness belt” marked so appropriately in such sombre black hues, and casting a deep shadow over our southern states which curls up to the economically depressed areas of the midwest, suggest something like a meaningful map, with the noting of neat exceptions of particularly happy cities, Asheville and Green Bay.   The weighing of these cities as exceptions lends a credence to the overall distribution of tweets the researchers collated in their data visualization, and the depth of data on which they relied.  The substantive study collated tweets over several years, even tracing computable variations in daily happiness averages that could be mapped to contemporary events, creating a set of stunning data visualizations in this “hedonometric” visualization from 2008 to the present whose units of days are suitably color-coded for weekday, allowing one to register how daily variations are effected by workdays and weekends.  The “hedonometer” seeks to provide the most accurately parsed chart of “happiness” based on daily counts of the tweeted words of happiness–the most common five words of happiness used each day suddenly appear when the day is hovered over.  The  graph is great fun to investigate, and can be tied to news events that impacted the nation’s overall index, from the Newtown shootings to inauguration days or holidays:  note the nation-wide spike on events like Christmas, which, since we still seem to all celebrate or at least note in some fashion, always reliable produces incalculable tweets.

If the first map from Time is a descendant and comedic successor to the UVM map of happiest states, both seem to rehabilitate the paper map in digital form as something like a response to the need for a “GPS for the soul,” an unfortunate mash-up if there ever was one.  Such maps exist in the big data-visualization echo chamber that has dominated our abilities to envision our country.  This echo chamber has existed ever since we came to believe that the country could be meaningfully cut up in meaningful ways for ready consumption.  If it could lie in the easy access to maps and data visualizations, it seems to respond to an unquenchable the thirst for images explaining regional differences that underly such a dichotomously divided status quo, since the division has roots that cannot be purely ideological in nature.

The single spectre that haunts the rise of even the most banal of data visualizations in media news in recent years may be maps of electoral results, especially from the Bush-Kerry 2004 election, in which that large red expanse of the middle of the country created a contrast to a close electoral contest of 296 to 242, which could have been upset by a single state.

Bush 296 - Kerry 242

The map haunted because it was almost repeated in 2008, with a key variation, only to be beaten back in recent years.

Obama:Biden McCain:Palin

These images seem to be seared into viewers’ minds, or at least into the unconscious of data visualizers.  Data of all sorts has since seeped into the map of the contiguous forty-eight.

Of course, the mother of all data-visualization maps is the most spectral, which still resonates with what some still consider the death-toll of democracy that at least one justice has come to regret:

 

ElectoralCollege2000

 

The contrast between that map and the popular vote led to something of a polemic exchange that was based on peering into data visualization maps to parse the vote, we might have forgotten, that familiarized everyone with data distributions:

 

County by County Bush v. Gore

 

The mapping of the country’s population has gained increased symbolic currency as a sort of transparent rendering of national opinions, only dreamed of in the early days of NORC’s General Social Survey, and far more easily visualized.  The creeping of data into such visualizations of the nation as “America’s Mood Map” has, after all, lent new authority to a visualization both more colorful and less depressing than the dichotomous division of the nation into “Red” and “Blue” states of almost Manichean terms.

And they are also much, much less depressing than the sort of heavy-handed Google Map divisions of the country into those regions that are ready to relinquish pre-K funding or subsidies, an idea that seems to undermine our national interest, as well as of those states that refuse the expansion of Medicaid, all in the name of undue federal influence.  To start with the first, we can view it two ways in news media, but both ways to illustrate the difficulty of ever arriving at consensus:  the below interactive (and informative) map that explores the educational opportunities in the Southern states of the US illuminates differences in pre-K funding (click on the above to explore funding changes in each state from 2009 to 2011, since the color-scheme is not self-evident).

 

SATELLITE VIEW-PRE-K FUNDING CITS

 

 

Below is a far more austere and stark way to visualize the data on how low many states rank kids less than four years of age, in which depression about care for pre-schools increases for the viewer in inverse relation to darkening of states’ hues.

PRE-K US 2005

In the colors of the data visualization blender, where data undermines map, there seems no consensus at all, and a pronounced fraying of the country’s diverse demographic.

One can always cut up the country in different ways, and the preferred way seems less based on splinters than blocks.  But some of the choropleths are striking and scary, as the refusal to expand health subsidies in the American Care Act, to which we’ll return.  The proliferation of these visualizations of difference may arise from the rise of the mythic “sea of red” in the general election of 2000 election through the Obama victory of 2008 may have left us barraged by the cutting up of the nation into camps.  The rise of new data visualizations seek to address these divides, but often seem to lie in the data visualization echo-chamber–as in the case of the “map of emotions”–as much as

But then there are those who reject either the Common Core standards or Affordable Care Act alike as forms of undue federal interference.

 

REJECTING COMMON CORE

 

Rejection of the ACA reveals a similar fragmentation, despite some serious number-crunching that went on to illustrate the high proportion of poor, uninsured and low wage-earning residents in may those very same blocks of states:

 

legend- Poor and Uninsured Americans

8% poor and uninsured

This is an odd echo, as I’ve elsewhere noted, between the very regions which outright refuse government expansion of Medicare and those with lack of insurance and large numbers of low-wage earners and some of the same states that refused to accept clearance by the Dept. of Justice before they changed voting procedures as an instance of undue federal interference.

 

Clearance Required

 

It’s nicer just to think that it all boils down to individual moods, which the scientific status of ““America’s Mood Map” nicely parse along clearly defined state-lines–even if its end results may have the scientific status of a mood-ring.  The chromatic variations are at least attractive, and able to be read easily, removed from political dissensus.  And it’s certainly more fun to imagine that we might be able to find a sense of constitutional differences inherent in the atmosphere of a region, and mirrored in lines of state sovereignty, that somehow miraculously reflect an almost Hippocratic sensibility of the shifting humoral constitutions of residents of different climates, rather than political or sociocultural (and socio-economic) differences.

But it’s hard to make any sense of the visualization, largely since the very values that it depicts do not lie on a continuum in the manner of most polls or degrees of gradual difference, but seem qualitatively distinct, and even, often, judgment calls.  The state-by-state map of personal constitutions hearkens back to an early modern notion of how place and season inform the humors, or regional climates color the mind.

It is perhaps not a far stretch to include a data visualization of a state-by-state map of obesity trends (and no doubt diet)–

 

OBESITY 2010–although such a map seems to isolate the deep south and its southern neighbors from Texarkana to New Mexico.

A vague overlap of data seems to exist similarly sized region, sadly, is plagued by lack of completing High School–although this has little relation to body-size, and there is little evidence of a relation between them, even if it does speak to the difficulty of valuing educational reforms like Common Core.

The difficulties created by “inadequate education” does seem to divide the country, however, as this choropleth reveals, and not only among those able to complete High School, but even in those who, having completed High School education, were not allowed to be part of the Army corps–a truly shocking statistic that effectively does divide the nation.

 

GRADUATION OF HS

 

Perhaps the only visualization that communicates unity is one of  cell-phone coverage, which customers, after all, desire–

 

Verizon-4G-LTE-Map-e1370794274644-540x327

 

By way of contrast, and a lightening of humors in how our country sees itself, “America’s Mood Map” shows a diversity around that one red block at its center, oddly located at Iowa–and whose deep red oddly seems to signify conventionality and friendliness–a quality the color does not suggest.

America's Mood Map

Other blocks of states are similarly lumped in oddly generic categories of states of mind–states of mind with limited relation to one another.  Hence, California, following, perhaps, conventional stereotype, is both open (if not that extroverted at all, particularly), and the among the least neurotic of the entire bunch.

 

Open and Un-neurotic California

 

In the most charitable reading permitted by the aggregation of data, the map would be an exercise in empathetic understanding of one’s neighbors limitations.   If one can permits an excursus, contrast it to the varied topography in the historic early modern “Carte de Tendre,” whose richly varied landscape suggests dangerous sites of delay or lack of clarity that the unaware and unsuspecting traveler may chance across by means of its locally detailed variations.

 

Carte de Tendre

 

These elegant enterprising travelers with cockades are gallant explorers of the outdoors, of course, rather than perched behind their screens.  Both the material and metaphorical nature of the data-visualization map are absent:  for in these cartographical transpositions, the data poses irreconcilable and absolute divides, and blocks any consensus from emerging.

“America’s Mood Map” is an artifact that serves as something of a mirror to make sense of our divided polity.  If one can given it a generous reading as an amusement, however, it may merit being taken seriously.  The eerily radical conceit of the data-visualization is not only that we are not “United” at all, but that one can naturalize states’ rights arguments in the radically different constitutions of their inhabitants, as if separate nations:  hence, conscientious Tennessee lies beside irascible Kentucky; open New York nearby to closed New Hampshire, and far from neurotic Maine; agreeable and conscientious North Carolina beside a Virginia that lags behind in both categories.  The authority that data is conceded in this visualization in fact erases mappable divides between rural and urban differences, socioeconomic distinctions, and patterns of wealth or any qualitative detail, taking the blocks of the electoral college as something like a national phrenological map.  The notion of an absolute difference in constitution as lying in direct relation to those state boundaries creates a particularly insidious illusion of differences that essentializes state lines–rather than following the idea of national character–that echoes one of the deepest presuppositions of what might be called Tea Party doctrine.  For the diversity depicted in data visualizations is always one engraved in hues of essentialization, rooting regions dispositions as fixed in a spectrum as different wavelengths, and empties the map of any continuity or local detail with those flat color blocks of distinctly defined individual “moods.”

How are you feeling?

Leave a comment

Filed under America, America's Mood Map, Care de tendre, choropleth maps, data visualizations, Google Maps, Hippocrates, Hippocratic humors, Jeffrey Kluger, MacPaint, mapping national divides, pre-K funding, Red states v. Blue States, Tea Party, Twitter map

Mapping Slavery in the United States in 1860

In debating the values of data visualization maps, I’ve gone both ways.  The value of maps as specific arguments–and tools of spatial orientation–respond to the value of the selective criteria that cartographers foreground in them, after all.  Less inventive differentiations of spatial distributions “flatten” the map’s surface, and limit their value to the map-reader.  Their arguments are not as interesting, one might say.  The early visualization of this elegant choropleth map employed data from the Census of 1860 it translated into visual form to map the density slave population across the recently seceded Southern states. Long touted as an important strategic tool, the rhetoric of an isolated mapping of the Southern states framed debates about the Civil War with greater  subtlety than current tiresome choropleth maps of “red” vs. “blue states.”  The 1861 lithograph marked the density of slave-owning in pockets by darkening sites of the greatest slave population, perhaps to mask the ownership of slaves throughout the South and point to the defenders of a slave-based economy.

1861 slave population map

If the census provided a basis for Edwin Hergesheimer and Alexander Dallas Bache to create the map, a collaborative government effort as much as an independent enterprise of the commercial engraver Henry S. Graham, the use of statistical cartography prefigured the mapping of social or political trend in the field of human geography.  While the recent German immigrant Hergesheimer created the pro-Union map from figures in the Census after his work on the US Coastal Survey, the translation of the results of the Census into visual form proceeded because the Coast Survey’s Superintendent.

Although a deeply collective project, Bache’s recent success in assembling a team to map the coastal survey gave him a new public profile, and prominence, to back the project of mapping a visual record of data assembled about slavery in Southern States.  The map tellingly reveals Hergesheimer’s deep Liberal opposition to slavery as an institution and the pro-Union belief of supervisors of the eighth national census, but its visual explanation of the origins of secession intentionally focussed attention around slavery debates.  Printed in September of 1861 after hostilities had begun, and ten states had seceded from the Union, it isolated the evil of slavery in the seceded region, and highlighted the centers of slavery’s institution even before slavery became the central issue of the war.

Hergesheimer and Bache were instrumental, too, in adopting the most current techniques of mapping to portray the differentials of slave-owning in the Confederate States just before the South’s Secession.  The half-tone engraving he designed to show Slave-Holding States in the Southern States of the United States in 1860 foregrounds discontinuities in the national territory by using figures he derived from that year’s national census.  Known as a “Slave Density Map,” the lithograph exemplifies cutting edge statistical mapping and an artistic use of half-tones to depict the seats of the evils of slavery in seceded regions of the United States in a piece of pro-Union propaganda:  slavery existed throughout the Southern states, but was concentrated “down river” in Mississippi, Georgia, South Carolina and Texas.  The lithograph provided something of a moral map of the region beyond which Lincoln sought to forbid the expansion of the slave-economy. Printed and sold by the government in wartime ostensibly for the benefit of Union Soldiers, the single-sheet map used half-tones to differentiate relative variations each county’s relative density of slave-holding across the Southern states.   But although the map presents itself as an appeal for wounded veterans of a war in which soldiers were so dramatically injured, it’s mapmaker aimed not only to raise funds through direct sales but broadly encourage the war effort by illustrating slavery as literally darkening the nation.

With considerable cartographical sophistication in charting variable densities of slave-ownership, the map illustrated striking discontinuities in the nation and  even suggested divisions in the Confederacy at the outset of the war:

Slavery Map 1860

Bache was a deeply moralistic man, as well as a former Lieutenant in the Army who, after graduating from West Point until 1829 designed coastal fortifications  before he rose to head the Smithsonian.  Bache had won national eminence as the head of the United States Coastal Survey, using a team of trained surveyors, several university observatories, and many field assistants to triangulate the coastlines from a base-line near Annapolis.  Based on readings from numerous sighting stations, the map extended west to California and endured through the twentieth century.  After mapping the shorelines of the United States, Bache devoted himself to mapping its islands of slavery.

Most maps made in wartime are valuable strategic tools to orient troops, as much as map vulnerabilities.  But this map was not made for a lack of cartographical records.  (It did not meet a desire for cartographical knowledge, for example, as the many paper maps shipped to Kuwait during the 1990 Gulf War–often incorrectly considered the first war fought by GPS–as 67 highly detailed topographic line maps of the region, some bathymetric, at the incredible scale of 1:50,000.   Incredibly, the Army lacked  accurate maps of Kuwait, and these were quickly prepared during the war based on remotely sensed satellite images.)  Bache’s map had far less apparent strategic use, identifying pockets where a slave-based economy was particularly dense, and all but ignoring physical topography or population centers.  But the map was compelling as an image of the divided nation.  Bache and Hergesheimer designed the map for a civilian populace, using the 1860 census to create a visually compelling distribution that revealed regional disparities in “slave-ownership” and the population identified as slaves to reduce the scope of the confederate secession.

The map provides an image of the nation taking stock of itself, or learning to look at its divisions–a rare thing.  Bache and Hergeshemier’s map of this nexus of the Southern economy was ostensibly sold to raise funds for “Sick and Wounded Soldiers of the US Army,” or veterans, as gothic script in scrollwork crisply notes.  It’s been argued that the map identified possible pockets of resistance, or possible seats of opposition to Union forces.  Lincoln consulted the map after the war, to consider ways to encourage the Southern economy.  But the elegant map roots the struggle of the Civil War in disparities of slave-ownership in the South, to question the fierceness of opposition to the Union, or to show the enemy and stakes in the war.  By starkly differentiating each county’s degree of reliance on slavery by shades of grey, it offered viewers a stark map of racism in parts of the region–if it also perhaps perpetuated racism  unwittingly in directing attention to the suffering of soldiers who could benefit from its sale and not sufferings inflicted by the economic institution of slavery.

legend

But the map also made its point of isolating slave-ownership in select regions of the Southern states, in ways that masked the continuity of slavery across the region.  The remove of the Southern states from the North, and isolation of precise regions where the practice of slavery was most extreme, echoed Bache’s belief in the political uses of science by applied cartography to national needs and for the public good.

It is interesting that Hergesheimer’s design for the map followed Bache’s success in precisely mapping the nation’s shorelines as Superintendent of the Coastal Survey.  The project established Bache’s credibility in large-scale surveying.  The new survey not only used the survey of coastal lines to map the interior, but may have provided Hergesheimer and Bache to turn attention to the occupants of the land as the US Government desired at that point.  The actual survey of coastal lines offered a sort of template to construct the map of slave-holding populations by using nine shades of half-tones to darken regions in differing degrees that created a compelling image of the fractured nation, and minimized the widespread nature of the social acceptance of slavery in seceding states.  The detail of the shore and coastal islands throughout suggests that the US provided not only contour lines for the map, but a template for national coherence–the unstated if implicit subject of the Slave-Holding map, and a central preoccupation of the wartime government.

The map of slave-holders adapted census information to the basic contours of states to map aggregate variations of slave-owning in states.  Although the map is given the strategic value of predicting resistance to the Union troops, it was probably most valuable for its inspirational or hortatory appeal as much as its accuracy.  Bache tried to expand the public functions of cartography in the map by adapting recent statistical methods to compellingly map two different worlds within the same nation.  A career military man and scientist, Bache was a public servant committed to the public utility and good of surveying and meteorology–a counter-part to Matthew Fontaine Maury, the polymath oceanographer and cartographer who served in the Confederate Navy 1861-5 and was Bache’s long-time nemesis:  Maury, whose cartographical interests I discussed in an earlier post, and who had himself hatched the ill-fated scheme for slave-owners to resettle from southern states to Brazil’s Amazon Valley in the early 1850s.  Bache’s map is both a detailed picture of social divisions and an image of a divided nation.  If this unlikely project was impractical, Bache’s map focussed on the seats of the slave-economy in seceding Southern states.

Susan Schulten noted in 2010, in a blog post to which I’m indebted, that Bache’s map shows a striking concentration ownership of slaves on the shores of the Mississippi, where cotton crops dominated, Alabama and the low-lands near South Carolina (which enslaved the majority of its population) and eastern Virginia; she suggested that this snapshot of slave-holding had strategic value to determine sources of greatest resistance in the Confederacy as well as for Lincoln to consider future economic development of the South.  The map must have constituted something like a “news map” for readers eager to understand the actual numbers of slave-owners in the deeper south, and the relative degree to which slave-owning had continued to endure in the Republic.  Parts of the broadsheet provided a strikingly compelling illustration of the locations of slave-ownership and an economy of slavery where Cotton was still King, as much as the Southern states as a whole:

Mississippi

Slavery was widespread in both Charleston, South Carolina, and Savannah:

Savannah and South Carolina

And, similarly, a pointed reliance on slavery around the area of Galveston, Texas:
Galveston

But there are multiple ways to read the project for designing the map.  The image of slave-holding lands provided a victory map for the Union soldiers, who sold it to raise money for their fellow-veterans who remained sick or injured, as if to stake the agenda for a need of remapping the nation, echoing a Jeffersonian idea of the use of surveying as a foundation for democracy.  It was the map by which Lincoln used to follow Union troops as they liberated slave populations, or understand the seats of rebellion, but had huge power in graphically stigmatizing Southern states.  It was also an idealistic statement of the goal of ending slavery in the war effort.

This striking legend that explained the iconography of the map’s nine variations in shading, placing the greatest “free blacks” in large towns:

Legend in Slavery Map

The innovative exercise in terrestrial cartography was also the last attempt to quantify slave-holding in the nineteenth-century, although it integrated public records that were later widely accessible.

It was, sadly, also the final time the Federal Government revisited the topography of slave-holding with similar precision. The failure to remap the same distribution seems one of the more stunning  cartographical silences of the twentieth century.  Bache’s impetus to draft the map might be informed by his long involvement in public education and belief in the public utility of the sciences; he effectively lent  a polemic character to conventions of statistical mapping by exploiting the different gradations of shading available to the engraver to craft a useful piece of early printed cartographical propaganda.

Indeed, its use as a piece of propaganda in wartime may have outstripped its potential as  guide to military strategy.  Lincoln regularly studied the map, according to his portraitist Francis Bicknell Carpenter, who found the president studying the map with considerable intensity in 1864, during the six months when he lived in the White House to paint the portrait “President Lincoln Reading the Emancipation Proclamation to His Cabinet.”  But his inclusion in the image of the choropleth map that accentuated national divisions reflected his interest in illustrating Lincoln’s “statesmanship” and “solid integrity” in bridging the nation.  By his own account, Lincoln had taken great pains to explain to Carpenter both the origins of “his adoption of the Emancipation policy” and his decision to draft the Proclamation from late July 1862, before he put it aside until being sure of military victory.  “I resolved,” Carpenter wrote in his memoirs, “to represent the scene [of Lincoln reading the Proclamation to his Cabinet] without the appliances and tricks of picture-making, and endeavor, as faithfully as possible, to represent the scene exactly as it transpired; room, furniture, and accessories all were to be painted from the actualities.”  Did the room actually include the map of slaves still ‘owned’ in the South?

Emancipation_proclamation-1

Carpenter’s state portrait depicting Lincoln reading the Proclamation to his Cabinet placed the lithograph in a small but a prominent role that most observers would not fail to notice.   As if to illustrate the subject of the recently issued Proclamation, if not the thoughts that weighed on the conscience of the President, the map emerged from behind a chair at the painting’s base.  Lincoln loved the portrait.  The map is the essential subject of his discourse; Carpenter made good use of it to capture the stakes of the Proclamation.  The nine bars of graduated shading in the map stand out among printed books in the group portrait, reminding viewers, both recording a moment of triumphalism and presidential dignity, and suggesting the uphill battle for implementing the Proclamation in the deepest South.  The map speaks volumes:

Visualizing Slavery-Carpenter

It reminds us not only of Lincoln’s hope to transcend social divisions.  Before Carpenter included it to illustrate the grand subject of Lincoln’s address–and Steven Spielberg also used the map in “Lincoln,” behind Daniel Day Lewis’ shoulder–the 1861 lithograph provided a tool to imagine political coherence in a more perfect union, urging men to enlist and others to get behind the war.

1 Comment

Filed under "Lincoln", Abraham Lincoln, Alexander Dallas Bache, choropleth maps, Edwin Hergesheimer, Matthew Fontaine Maury, US Coastal Survey, West Point