Let’s be honest—there is no def­i­n­i­tion of dig­i­tal human­i­ties, if by def­i­n­i­tion we mean a con­sis­tent set of the­o­ret­i­cal con­cerns and research meth­ods that might be aligned with a giv­en dis­ci­pline, whether one of the estab­lished fields or an emerg­ing, trans­dis­ci­pli­nary one. The cat­e­go­ry denotes no set of wide­ly shared com­pu­ta­tion­al meth­ods that con­tribute to the work of inter­pre­ta­tion, no agreed upon norms or received gen­res for dig­i­tal pub­li­ca­tion, no broad con­sen­sus on whether dig­i­tal work, how­ev­er defined, counts as gen­uine aca­d­e­m­ic work. Instead of a def­i­n­i­tion, we have a geneal­o­gy, a net­work of fam­i­ly resem­blances among pro­vi­sion­al schools of thought, method­olog­i­cal inter­ests, and pre­ferred tools, a his­to­ry of peo­ple who have cho­sen to call them­selves dig­i­tal human­ists and who in the process of try­ing to define the term are cre­at­ing that def­i­n­i­tion. How else to char­ac­ter­ize the mean­ing of an expres­sion that has near­ly as many def­i­n­i­tions as affil­i­ates? It is a social cat­e­go­ry, not an onto­log­i­cal one.

As a social cat­e­go­ry, the term has a more or less clear set of orga­ni­za­tion­al ref­er­ents. Recent­ly Matt Kirschen­baum remind­ed us that there is a peer-reviewed jour­nal, a fed­er­al office, an annu­al con­fer­ence, and an inter­na­tion­al net­work of aca­d­e­m­ic cen­ters asso­ci­at­ed with the term, not to men­tion an Oxford Com­pan­ion. How­ev­er the gap between the social and the onto­log­i­cal can­not avoid appear­ing as a kind of scan­dal. This is evi­dent from the num­ber of essays and blog posts that have emerged seek­ing to define the cat­e­go­ry, as well as from the play­ful­ly com­bat­ive and defen­sive tone some remarks have tak­en. This anx­i­ety of self-def­i­n­i­tion seems to indi­cate a new phase in the his­to­ry of the field, one that may indi­cate the emer­gence of a ter­ri­to­r­i­al instinct in an envi­ron­ment of scare resources–even as the lan­guage of the “big tent” emerges. After all, the shift from Human­i­ties Com­put­ing to the Dig­i­tal Humar­ni­ties index­es a growth in the size and pop­u­lar­i­ty of the com­mu­ni­ty. With growth comes grow­ing pains.

To many, the dig­i­tal human­i­ties feels like a small town that has recent­ly been rat­ed as a great place to raise a fam­i­ly. It is now inun­dat­ed by devel­op­ers who want to build con­dos for new­com­ers who are com­pet­ing for resources and who may not under­stand local cus­toms. Iden­ti­ty crises emerge when tac­it, unspo­ken under­stand­ings and modes of inter­ac­tion are dis­rupt­ed by exter­nal con­tact and demo­graph­ic shifts. In the quest to defend old ways and invent new ones, in-groups are defined, prophets emerge, witch­craft accu­sa­tions are made, and peo­ple gen­er­al­ly lose what com­mu­nal sol­i­dar­i­ty they once had. The dig­i­tal human­i­ties com­mu­ni­ty has not gone this far, but one can­not help but notice the dis­par­i­ty between the Wood­stock feel­ing of THAT­Camp events and what appears to be the Alta­mont of DH 2011.

To be sure, all dig­i­tal human­ists share a com­mon bond as human­ists, schol­ars devot­ed to the inter­pre­ta­tion of what Panof­sky called “the records left my man [sic]”—works of lit­er­a­ture, art, archi­tec­ture, and oth­er prod­ucts and traces of human intel­lec­tu­al labor. More specif­i­cal­ly, the sorts of human­ists who have been drawn into the fold of dig­i­tal human­i­ties have had a dis­tinct pref­er­ence toward tex­tu­al remains, even if we enter­tain pleas to con­sid­er non-ver­bal chan­nels as well (usu­al­ly orig­i­nat­ing from non-tra­di­tion­al fields, such as media stud­ies). It remains an implic­it (if dis­com­fit­ing) assump­tion among dig­i­tal human­ists that, as Tim Bray put it, “knowl­edge is a text based appli­ca­tion.” Con­sis­tent with this view, the typ­i­cal dig­i­tal human­ist is a lit­er­ary schol­ar, an his­to­ri­an, or a librarian—all tra­di­tion­al fields con­cerned with the man­age­ment and inter­pre­ta­tion of writ­ten doc­u­ments. Oth­ers, such as myself, come from oth­er back­grounds; but I believe it is no acci­dent that the most recent buzz about the dis­ci­pline was spawned by talks giv­en at the MLA.

There are also many schools of thought under the sign who do share, with­in them­selves, a more or less coher­ent set of meth­ods and con­cerns. There is, of course, the old guard of human­i­ties com­put­ing, trained in the dig­i­ti­za­tion of tex­tu­al sources using TEI and versed in the the­o­ret­i­cal impli­ca­tions of this mode of rep­re­sen­ta­tion. There is a new­er com­mu­ni­ty who embrace the “spa­tial human­i­ties” through the use of map­ping soft­ware in rela­tion to tex­tu­al (and oth­er) sources, and who have shift­ed our atten­tion toward visu­al­iza­tion and human geography—an over­looked field that should right­ly have its day. Along­side these there is a long run­ning group of sta­tis­ti­cal crit­ics, extend­ing from Father Busa and IBM to Fran­co Moret­ti and Google, as well as oth­er com­pu­ta­tion­al human­ists who have been at it since the 1960s and who believe that count­ing words, apply­ing the meth­ods of com­pu­ta­tion­al lin­guis­tics, and observ­ing pat­terns in large cor­po­ra will pro­duce insights unreach­able by mere read­ing. One could also point to the Crit­i­cal Code Stud­ies group and oth­er schools of thought that have emerged in the space.

Tak­en as a whole, how­ev­er, there is lit­tle con­nec­tion among these groups beyond a shared inter­est in texts and the use of com­pu­ta­tion­al tech­nolo­gies to explore and under­stand them (as opposed to mere­ly cre­at­ing or dis­trib­ut­ing them). But more impor­tant, none of these groups, either in iso­la­tion or as a whole, has suc­cess­ful­ly demon­strat­ed to the wider com­mu­ni­ty of human­ists that there are essen­tial and irre­place­able gains to be had by the appli­ca­tion of dig­i­tal tools to the project of inter­pret­ing (and rein­ter­pret­ing) the human record for the edi­fi­ca­tion of soci­ety. To a dis­con­cert­ing­ly large num­ber of out­siders, the dig­i­tal human­i­ties qua human­i­ties remains inter­est­ing but irrel­e­vant. Antho­ny Grafton speaks for the major­i­ty when, in a recent New York Times piece, he repeats the plat­i­tude that the dig­i­tal human­i­ties is a means and not an end. (Giv­en his stature in the field, his recent Road to Dam­as­cus expe­ri­ence at the AHA may indi­cate a turn­ing of the tide—but the con­ver­sion of oth­er promi­nent schol­ars has not pro­duced such shifts in the past.)

Now, if we use the term dig­i­tal human­i­ties and can­not define it, maybe we are think­ing of such def­i­n­i­tions in the wrong way. Maybe the tra­di­tion­al way of defin­ing dis­ci­plines in the acad­e­my is all wrong. Instead of say­ing that physics is the study of mat­ter and ener­gy, or his­to­ry the study of what peo­ple have done in the past, maybe we should say that physics is the work of those who read New­ton and Ein­stein, who use var­i­ous branch­es of math­e­mat­ics, and who know how to con­struct exper­i­ments in a cer­tain way. Or his­to­ry is the work of peo­ple who know how to nav­i­gate archives and read old tax records and diaries and oth­er tex­tu­al remains, where­as archae­ol­o­gists are those who know how to man­age digs and how to retrieve, clas­si­fy and inter­pret shards and bones.

This may sound forced for the hard sci­ences, but it is emi­nent­ly rea­son­able for the human­i­ties and social sci­ences. For what are the real dif­fer­ences between his­to­ry, soci­ol­o­gy, eco­nom­ics, anthro­pol­o­gy, and archae­ol­o­gy? Each claims to address the struc­ture and func­tion of soci­ety. The answer is that each has mas­tered a par­tic­u­lar domain of data—its acqui­si­tion, orga­ni­za­tion, analy­sis, and inter­pre­ta­tion. Soci­ol­o­gists do sur­veys and sta­tis­tics, inter­views and con­tent analy­sis. Cul­tur­al anthro­pol­o­gists do field­work and thick descrip­tion. Econ­o­mists count indi­ca­tors and devel­op equa­tions to relate them. His­to­ri­ans are very good at con­vert­ing old doc­u­ments and archives into sto­ries. When an archae­ol­o­gist starts to read such doc­u­ments, we say she is doing “his­tor­i­cal archae­ol­o­gy.” Doc­u­ment-read­ing anthro­pol­o­gists become eth­no­his­to­ri­ans. And so forth.

Such a def­i­n­i­tion (which philoso­phers will rec­og­nize as a species of prag­ma­tism) allows us to turn our atten­tion to the prac­ti­cal and sit­u­at­ed basis of the dig­i­tal human­i­ties. In this view, dig­i­tal human­ists are sim­ply human­ists (or inter­pre­tive social sci­en­tists) by train­ing who have embraced dig­i­tal media and who have a more or less deep con­vic­tion that dig­i­tal media can play a cru­cial, indeed trans­for­ma­tive, role in the work of inter­pre­ta­tion, broad­ly con­ceived. Beyond this all bets are off. Because the cat­e­go­ry of dig­i­tal media includes essen­tial­ly every­thing afford­ed to the human­ist by the pres­ence of avail­able computing—everything from crowd-sourc­ing and social media to nat­ur­al lan­guage pro­cess­ing and latent seman­tic index­ing to gam­ing and hap­tic immersion—the dig­i­tal human­i­ties is in prin­ci­ple asso­ci­at­ed with as many meth­ods and tools as there are inter­sec­tions between texts and technologies.

The com­plex­i­ty of the field is also mul­ti­plied by the modes of rela­tion­ship that may char­ac­ter­ize the inter­sec­tion between com­pu­ta­tion and tex­tu­al­i­ty in each case. Con­sid­er the dif­fer­ence between the prac­tices of tex­tu­al markup and the work asso­ci­at­ed with Crit­i­cal Code Stud­ies. The for­mer sub­jects pri­ma­ry source texts to dig­i­tal rep­re­sen­ta­tion by means of code—XML, XSLT, etc.—whereas the lat­ter treats code itself as text, seek­ing to apply prin­ci­ples of inter­pre­ta­tion theory—hermeneutics, struc­tural­ism, etc.—to pro­gram­ming lan­guages and, one hopes, markup lan­guages as well. (One might include here Kirschenbaum’s Mech­a­nisms, which treats hard­ware itself as text.) As Stephen Ram­sey points out some­where, prac­ti­tion­ers of the for­mer approach can be curi­ous­ly uncrit­i­cal of their tools and meth­ods, check­ing their post­mod­ernist per­spec­tives at the door of the lab.

Con­sid­er also the case of data­bas­es. On the one hand, many schol­ars sup­ple­ment their research by using data man­age­ment tools to orga­nize notes and ref­er­ences. On the oth­er hand, there is an emerg­ing school of thought, ini­ti­at­ed by Lev Manovich, which regards the data­base itself as an object of crit­i­cism in its own right. The dif­fer­ence between the two approach­es is like night and day, although one can imag­ine how one may prof­it from the oth­er. Still a third mode of inter­sec­tion is to regard tech­nol­o­gy as an alle­go­ry of tex­tu­al­i­ty. For exam­ple, Wendy Hui Kyong Chun has employed the image of the fiber optic net­work as a frame for the inter­pre­ta­tion of dig­i­tal­ly medi­at­ed social inter­ac­tion and text. So not only are there as many kinds of dig­i­tal human­i­ties as there are inter­sec­tions between human­i­ties and com­pu­ta­tion tech­nol­o­gy, that num­ber is at least tripled by the kind of rela­tion­ship that inheres in that inter­sec­tion. To a human­ist, com­pu­ta­tion­al tech­nol­o­gy is poten­tial­ly tool, text, and metaphor.

Giv­en this sur­plus of exten­sion­al mean­ings, there is sim­ply no way to describe the dig­i­tal human­i­ties as any­thing like a dis­ci­pline. Just think of the cur­ric­u­lar require­ments of such a field! Not only would the field require its mem­bers to devel­op the deep domain knowl­edge of the tra­di­tion­al humanist—distant read­ing notwithstanding—it would also demand that they learn a wide range of diver­gent tech­nolo­gies (includ­ing pro­gram­ming lan­guages) as well as the crit­i­cal dis­cours­es to sit­u­ate these tech­nolo­gies as texts, cul­tur­al arti­facts par­tic­i­pat­ing in the repro­duc­tion of social and cog­ni­tive struc­tures. Grant­ed the occa­sion­al poly­math who may mas­ter all three, the scope of such a pro­gram is sim­ply too vast and var­ie­gat­ed. And in fact there has been no con­sen­sus among dig­i­tal human­ists about the basic ele­ments of a cur­ricu­lum, a prob­lem we share with advo­cates of media flu­en­cy to define a cur­ricu­lum for fac­ul­ty development.

So, if the dig­i­tal human­i­ties is, nei­ther in fact nor in prin­ci­ple, a dis­ci­pline, then what is it? Sure­ly, with its grow­ing army of fol­low­ers and pletho­ra of con­crete insti­tu­tion­al man­i­fes­ta­tions, it must have some basis in a real­i­ty oth­er than its own exis­tence. In fact it does. The dig­i­tal human­i­ties, as both a broad col­lec­tion of prac­tices and an intense, on-going inter­pre­tive prax­is gen­er­a­tive of such prac­tices, is best thought of as hav­ing two very con­crete but equal­ly elu­sive dimen­sions. One the one had, the dig­i­tal human­i­ties (con­ceived of in the plur­al) com­pris­es some­thing very much like a cur­ricu­lum, an inter-relat­ed col­lec­tion of sub­ject domains and resources that, as a whole, con­tributes to both the con­struc­tion of knowl­edge and the edu­ca­tion of peo­ple. Although no one indi­vid­ual can mas­ter an entire cur­ricu­lum, a cur­ricu­lum nev­er­the­less has a log­ic, a coher­ence, and even a cen­ter of gravity.

This leads to the sec­ond and more impor­tant dimen­sion: that cen­ter of grav­i­ty is not a par­tic­u­lar assem­blage of tech­nolo­gies or meth­ods but the on-going, play­ful encounter with dig­i­tal rep­re­sen­ta­tion itself. It is the encounter that the dig­i­tal human­ist dis­cov­ers and finds at once reveal­ing, sat­is­fy­ing, and an inef­fa­ble source of fel­low feel­ing with his col­leagues. This encounter is not regard­ed as mere­ly a means to an end, but as an end in itself, in so much as the process of inter­pre­ta­tion is often as reward­ing as its prod­ucts. I call this encounter the sit­u­a­tion of dig­i­tal rep­re­sen­ta­tion, a sta­ble but always-in-flux event space that is but a spe­cial case of the work, or prax­is, of rep­re­sen­ta­tion in gen­er­al. Adult mem­bers of lit­er­ate cul­tures for the most part have sub­li­mat­ed and for­got­ten this prax­is, but it remains present to the minds chil­dren and poets, who are always learn­ing how to read and write.

This, I believe, is what Stephen Ram­sey means by “build­ing.” Or at least it is a char­i­ta­ble mis­read­ing that retrieves the argu­ment he makes when he sug­gests, essen­tial­ly, that real dig­i­tal human­ists write code. In my rephras­ing, real dig­i­tal human­ists are engaged in the play of rep­re­sen­ta­tion, which pro­found­ly involves putting things togeth­er, whether the vehi­cle of assem­bly be Lisp or Zotero. That marks a wide spectrum—but there is a com­mon ele­ment of play, of pro­duc­tive­ly map­ping and remap­ping the objects and cat­e­gories of schol­ar­ship in the rapid­ly chang­ing, intrin­si­cal­ly plas­tic but struc­tural­ly con­strain­ing media of dig­i­tal tech­nol­o­gy. With­out this play—to the extent that the schol­ar has a stand-off, do-this-for-me atti­tude toward the medium—then, no, she is not a dig­i­tal humanist.

Dig­i­tal human­ists are aware that in the cur­rent his­tor­i­cal moment, as the old­er men­tal­ites of print lit­er­a­cy con­tin­ue to be dis­placed and reworked, the human­ist has the oppor­tu­ni­ty to immerse her­self in the trans­duc­tive plas­ma of inter­pre­ta­tion where ideas and their expres­sive vehi­cles can be mapped and remapped in a vari­ety of forms and frame­works, a gid­dy play of prax­is that not all gen­er­a­tions have the good for­tune of wit­ness­ing. This expe­ri­ence cross-cuts all of the var­i­ous dis­ci­pline- and tech­nol­o­gy-spe­cif­ic instances of dig­i­tal human­i­ties work. To the extent that a com­mon dis­course is emerg­ing to reflect on this expe­ri­ence across the dis­ci­plines, then the dig­i­tal human­i­ties is a real enough.