(links and images still need to prettify’d, but the text is set)
(|ARCHIVE|) The final installment in a series of six readings and interpretations from The Geography of Beer (https://www.amazon.com/Geography-Beer-Regions-Environment-Societies/dp/9400777868) , Part III - Societies. The abstract of the current paper, ‘Offline Brews and Online Views: Exploring the Geography of Beer Tweets’ by Matthew Zook and Ate Poorthuis (mailto:firstname.lastname@example.org?subject=Offline%20Brews%20and%20Online%20Views%20%7C%20EndlessPint%20Write-up%20) , may be read on Springer (https://rd.springer.com/chapter/10.1007/978-94-007-7787-3_13) .
All I see are blonde, bruin, stouts, IPA…
Offline Brews and Online Views is a wonderful illustration of a magician’s trick. The authors have a twofold motivation: to map the spatial occurrences, frequencies, and comparative differences of Tweets related to beer and to connect this online representation with offline consumer behavior. These motives are complimentary and helpful in informing one another. They are also separable and remain so, despite the numerous incantations of the authors. Did I say magicians? I meant hypnotists.
The authors reference the online presence of beer mentions and their close ties and relations to offline behavior. Yet beyond the Tweets collected, counted, and mapped there is no direct tie-in to actions in the real world. The closest reference to the real world is in identifying wine producing and traditional brewing regions and their underpinning to the Tweet activity. We are reassured that we are seeing what we should expect, but there is no data link between the Tweets, their contents, geolocation, and specific beer sales. All we can say with certainty is that the maps represent the typing of Global Teens (https://www.tor.com/2011/04/12/genre-in-the-mainstream-the-dystopia-of-gary-shteyngarts-super-sad-true-love-story/) on their keypads on one social media platform.
The maps are representative of the data set collected. If that sounds circular, it is. The data visualizations are self referential and need not reflect any further underlying reality. The data are complete because they are self contained. The authors would have us believe that they are showing something more significant, and that is open to debate, but claiming “the complex intertwining of offline preferences”, “material geographies”, “offline material presences”, “how tightly imbricated the material and digital worlds are” does not make it so. Unless we forget about what is not represented and begin taking the map for the territory. At that point we can just as easily reverse the order and wonder at how offline behavior reflects online preferences.
To the extent that we can agree on the appropriateness of the data targeted, the correctness of its collection, the soundness of the processing, sorting, and sampling we can unanimously say that the generated maps are successful in representing the Tweets of beer users over a twelve month span in the contiguous United States. The fidelity to the source data is not in question but remaining locked in on this aspect of the analysis helps float the smoke screen on its disassociation with offline behavior. The beer representations and findings fall in line with expectations, so the authors keep reminding us, but there is no direct tie-in to actual sales, except, if we are being generous, in the instance of providing an ordinal listing of the top light lagers sold over the given time period, though the authors “do not make [a] direct connection between Tweets and sales, as the latter data are not available.”
There is no duplicitousness in the paper, the authors point out where their analysis stops short. Besides sales they also do not take into consideration the sentiments of the Tweets and they admit to certain geographic areas being left out for lack of data. The authors are honest in their presentations but it appears that they have enchanted themselves into believing the maps show more than they do. Constant allusions to offline behavior are made but no evidence is provided, aside from references to wine producing and traditional brewing areas. As experts the authors may have good reasons for their claims but within the evidence provided the assertions remain unfounded. There is no connection to sales, tax data, beer distribution, demographic analysis or discussion of income and drinking preferences.
There is the reality of beer sales and consumption, what that means for and about our society, regionally as well as in total. When it comes to sales we can gather these numbers in macro, across the country, or in micro, by region and brewery. The online behavior of consumers can serve as a gauge but it is a proxy. Presenting a map of beer activity and attention as a function of Tweets with no underlying tie-back to real world actions results in a disconnect. The data and resulting maps are self referentially accurate, true to what they measure and display, Tweet activity, but they are unmoored to the reality underneath. Suggesting the maps as representations of reality is probably more true than the authors intended but it is a truth about the data and the activity pertaining to its creation not the truth of offline behavior, especially so long as the sales, taxes, etc. are left out of it.
https://en.wikipedia.org/wiki/On_Exactitude_in_Science J. L. Borges The resulting maps are a simulacra of beer activity. They are not copies, they are not abstractions. There is no concealment of truth, the authors are forthright in the limitations of their approach, but the map conceals, perhaps to the authors as well, that there is no truth, not beyond the digital representation. It is true to the online behavior of users, a reality separate from the consumption of beer. Of course, Tweets happen alongside beer activity so there is contemporaneity about the data and the beer, simultaneous existences, but “[a]ny naive pretension to reality as such is perceived as bereft of critical self-awareness” which may be in part what our authors are guilty of by constantly imploring us that the maps overlap with reality.
The hypnotic chantings regarding the map and reality matching obscures how much we are asked to take this on faith. Eventually we may rely solely on the map. Not just as a proxy but as a fair representation of reality. Though the uptick or decline in Tweets would mean nothing without a correlation to sales, a mention to marketing campaigns, and seasonal promotions (sports, holidays, etc), simply registering the Tweets, their trend, and changes over time becomes a measure on its own. Eventually this “simulacrum precedes the original and the distinction between reality and representation vanishes.” The data is available, we can collect and measure it. Presenting it is true, but true of itself. It is a self contained and self referential truth. Soon we speak of the map and its changes no longer referencing the underlying economic and social behavior and in time references to the map become part of the stories we tell one another. “The map that precedes the territory - precession of simulacra - .. is the map that engenders the territory.”