back to

Dan Hoy

The Virtual Dependency of the Post-Avant and the Problematics of Flarf: What Happens when Poets Spend Too Much Time Fucking Around on the Internet

This piece is 11,013 words
or about 13 printed pages long.

Google is not a spontaneous manifestation of the zeitgeist in the virtual realm. That it is often misconstrued as such is due to a passive acceptance of its process and mythos, from its humble beginnings and benign-sounding name to the embedded cultural belief that the Internet is the great democratic frontier in which all information is equalized — the user, instead of the disseminator, is the arbiter of what is useful and not useful — and the residual PR advantage this gives to in utero virtual corporations like Google and Yahoo! over preexisting technocapitalist transplants like Microsoft. Google is considered an organic entity only marginally different from a construct like Wikipedia, an open-source encyclopedia in which users define the content by continually creating, altering, contesting, and amalgamating entries.[1]

But there’s a significant difference between the technophile’s ideal of what the Internet can be and the tools in place to traverse what it is. The Internet may be rhizomatic — but search engines are not. They’re selectively hierarchical. That poets are employing these hierarchies as poetic tools without questioning the implications of doing so (whether in pre- or post-production) exposes a lack of rigor in their process, as well as a tacit disregard for their own cultural complicity as something maybe worth exploring, or at least being aware of. And that many of them go on to talk in venerating tones about search engine collage (often referred to as ‘flarf’ or ‘Google-sculpting’) as a catalyst for poetic enlightenment as well as a revolution in poetic technique suggests a quotidian misreading of recent artopoetic history, most prominently embodied in John Cage and the various members of Oulipo.

Gary Sullivan (on his blog E l s e w h e r e) makes the claim that K. Silem Mohammad’s Deer Head Nation ‘turns Gutenberg’s handheld wonder into a pinball machine.’[2] It’s not my intention to dispute or agree with this kind of assessment of Mohammad’s poems and other flarf. In fact it’s not my intention to read them as poems at all, but as collage. And even then, the focus of this essay is not a reading of flarf per se but a reading of the reading of flarf (the critical reception of their poems as well as the poets’ stated intentions). My primary source then is what most people would consider an auxiliary source, but in the context of this essay the poems themselves are beside the point. In any case, I think the way these poems are written and received is very much a part of their content; the former because it informs them as they’re developed (inherent content), the latter because it determines their functionality (projected content). Though you could tell me that all content is projected and I wouldn’t disagree with you — but that only affirms my position that what is commonly defined as extratextual is an ideological consideration, not a fact. On a similar note, one could argue that all poetry, that all speech even, is collage, since it’s an assemblage of words appropriated from different sources (e.g. parents, kindergarten, street signs, cereal boxes, TV).[3] But by collage I mean an artwork whose appropriation of material is part of its material, for example

awe yea I open a photo album I found under my bed
uhhuh, The dusty, leather cover decaying and smelling of the years
awe yea baby Regrets mingling with my tears
as I methodically turn the pages, you see
I like to dress up in REALLY tight underwater pumpkin beavers...
and I take a deep, painful breath
Because staring back at me from the tattered oragami licences
oh baby yea Are black and white visions of faraway hearts uh huh

from Drew Gardner’s ‘As Dolphins Languor’. (Though this example is not Google-assisted flarf: see note [4]). As collage, Google-sculpting is necessarily polyvocal on its surface. To praise or dismiss it for this is to miss the point. What’s important is that we read it with its implicit questions in mind. How is it assembled? From what is it assembled? What is the relationship between this new amalgamated context and the old unrelated contexts? It’s this type of reading, not its classification as ‘poetry’ in relation to other ‘poetry’, that has been inadequately addressed in formulations of flarf and the exegesis of it.

I should point out there’s a more general definition of ‘flarf’ that’s not dependent on corporate algorithms, whose criteria is simply that it be ‘a blend of the offensive, the sentimental, and the infantile,’ (K. Silem Mohammad) [5] or, even more simply, ‘A kind of corrosive, cute, or cloying, awfulness. Wrong. Un-P.C. Out of control. “Not okay.” ‘ (Gary Sullivan).[6] In fact Mohammad argues on the 07.07.05 entry for his blog {lime tree} that Deer Head Nation isn’t flarf at all — ‘I’m trying desperately to avoid calling it “post-flarf,” but I may yet cave in’, which he does — maintaining that the introduction of ‘concentrated Google collage methods’ into flarf actually precipitated flarf’s shift into ‘post-flarf’.[7] But I’m not interested in getting into a semantic argument about it, especially with a poet who first declares ‘There is no such thing as Flarf’[8] only to later ask ‘Why did flarf die?’[9] and who’s also quoted himself as saying ‘Flarf is intentionally bad poetry that involves Google search text results’ to a BBC interviewer.[10] I’m employing flarf here as a term that’s used by bloggers in reference to ongoing methodologies, including that of fellow flarfist Jordan Davis, who responded to Mohammad’s evasion with ‘But as a member of the shadow cabinet of the party of Flarf, I have to object. I object, sir! Sir! Your boxer-brief-button is broken and your bling bling pachydeerm is Joni Mitchelling on the sidewalk!’[11] Regardless, this essay is not about that kind of flarf.[12]

But this paragraph is (and so is the next one). And Mike Magee’s flarfy aspirations are more directly political than his fellow Un-P.C.s-in-arms, as evidenced by his prefacing a discussion of poetics with a W.C. Williams quote about poetry in relation to war, as well as his following this contextualization with ‘I have in mind, always now, Frederick Douglass’s words, “At a time like this, scorching irony, not convincing argument is needed”’, which scorching irony Magee translates as straight-up camp in the form of a ‘Fascist Fairytale’ starring Margaret Thatcher and the Sphinx that ends with Thatcher exclaiming ‘Bomb Turks, I’m in love!’[13] But one has to wonder about the efficacy of this response when Magee himself, three paragraphs earlier, observes that ‘The state has always attempted to co-opt the language of dissent and so de-fang it’. And the problem with Magee’s version of ‘scorching irony’ is that it’s been incorporated into television commercials and other corporate/ state outlets for some time now, as discussed at length in David Foster Wallace’s diagnostic essay ‘E Unibus Pluram: Television and U.S. Fiction’: satiric irony as a form of THIS IS SATIRE cultural critique just doesn’t cut it when the thing being critiqued is already an ironic satire of itself, one that in fact invites you to read it as such so as to induce a state of self-congratulatory knowingness as you and the thing share a wink-wink nudge-nudge inside joke of Escherian proportions.[14] Though I think Magee’s appropriation of the phrase ‘Mainstream Poetry’ for a movement and domain name is pretty funny.

In any case, flarf’s Un-P.C. aesthetic aim (I’m still talking about the more general flarf) is a reactionary sentiment left over from the 90s and indicative of what’s problematic about the virtual poetry community’s closed circle of interlinked blogs (with Ron Silliman as the most interconnected and therefore central hub[15]). A juvenile strategy like this may be revolutionary in relation to the Mary Olivers[16] of the world, but if one follows Duchamp by dropping the medium as a defining limit so as to engage with poetics as a behavior and way of thinking, it’s not a matter of creating poetry in relation to other poems: ‘poetry’ is simply a byproduct, and what leads to these recursive acts of ‘revolution’ is a defining/ confining of it to the rubric of letters and lettres, text and txt. If Mohammad’s Great Idea is that poetry should be ‘as shameless as television’,[17] why doesn’t he just drop the ‘poetry’ and go intern at Fox News or Spike TV or some place well-funded/ versed in the saturating effects of an all-out audio-visual and ideological assault on one’s audience? I’m not kidding. Television already does what flarf does, but with a greater agility and honesty than the flarfists since it’s had decades of experience with target audience research: not only is it aware of the absurdity of its cultural power, it manipulates that absurdity as a means to its marketing ends. And how are flarf’s aspirations, with its poetics of awfulness as a style, any different than say Jimmy Kimmel Live?[18] As long as poets limit their poetic output to text (which is fine in itself) while also limiting their conception of ‘poetry’ to it, we’re going to keep getting these episodic moonings and middle fingers made against some stuffy fuddy-duddy or otherwise projected entity (in this case ‘acceptable’ poems and poets) under the guise of a critical leveling of the larger culturopolitical establishment. As if any boundaries are transgressed when you hold a protest within the space (material or cognitive) apportioned off for you by the institution you’re protesting against. This is inversely related to discussions about poetry’s marginality in pop culture and the arts (poetry is the big game / poetry is the sidelines), and it’s exactly the kind of myopic view of what it is to engage with the world (in)directly as the greatest poem (like Whitman’s America) that makes me think poets have learned nothing from the 20th Century they didn’t already know going in. Obviously this isn’t the case, but it’s what it makes me think.

But the rest of this essay is about the uncritical use of corporate algorithms as a generator of poetic chance and catalyst for engaging the Other, a trend[19] among the ‘post-avant’ or whatever you want to call it (including poets whose work is otherwise interesting) that betrays not just their mediated upbringings but an antiquated technophilia. The proliferation of flarf and its hybrids recycles an industrial era excitement over human ‘progress’ with no hesitance toward the embarrassing hubris of such a perspective. It’s retro-Futurist, and it’s indicative of their reliance on the virtual realm as a method of navigating reality. Not to dispute the body-extension or trans-geo and -temporal benefits of virtual technology — but these benefits also imprint the behavioral belief that one can do everything by doing nothing. And what’s happening is a willful dependency on corporate tools to do the searching, selecting, and contextualizing of poetic material, with no intra-textual suspicion or extra-textual analysis of the tool itself or what this means for the ‘product’ that’s being made. A case in point is the Get a Google Poem website, which eliminates the middleman (you) altogether by producing a seemingly instantaneous readymade Google-sculpted poem for the user based on a checklist of parameters.[20] The site is run by poet/ translator/ novelist Leevi Lehto, whose fanboy complicity with Google is on display (literally, on the site) in a letter he wrote to ‘the guys [sic] at the world’s favorite search engine’, in which he defined his website as ‘a collective artistic celebration to, and reflection on, the great revolution in the structures of information delivery and retrieval that Google is so much [a] part of’.[21] I’m assuming he means ‘reflection of’ instead of ‘reflection on’: with nothing but celebratory perpetuation on the site, the ‘reflection’ is a virtual mirror, not a critical analysis.

Charles Bernstein may call Deer Head Nation a ‘necessary ethical parry’[22], but any claim of flarf as sociopoetic engagement is undermined by not including the search engine (which is talking as loudly as the voices it PageRanks) as part of the social network. This isn’t to say I don’t find some of the poems in Deer Head Nation interesting when decontextualized (bypass the trying-too-hard ‘Mars is for Terrorists’ Lyn Hejinian picked for Best of American Poetry 2004 and other show-offy dick-wagging for quieter, odd moments like ‘World’s Largest Robot Store’), or that the flarfists, in general, are as awful as they feign to be (check out for example David Larsen and Maria Damon in Combo 12 (edited by Mike Magee), the unofficial ‘Flarf Issue’, or many of the Katie Degentesh poems available online and elsewhere). Nor is the interconnective insularity of blogs and listservs an inherently good or bad thing for the poets who use them; but in the context of a collective gung-ho attitude regarding their own techno-dependency, it’s a breeding ground for some bad habits and a utopian view of the Internet’s impact on the poetry community (which is often conflated with poetics). Take this example from the 09.28.05 entry on The Unquiet Grave, the blog for Tony Tost (a poet whose work I admire), who’s been an unequivocal proponent of flarf through Grave and his online journal Fascicle (a journal whose content and aspirations I admire):

imagine I also wrote a couple of paragraphs on why Flarf is interesting to me because: Andre Leroi-Gourhan argues that technics are externalization of internal processes, Bernard Stiegler revises him to say that the technical (external) and internal co-evolve, and that I think Flarf is one of the first instances I see of this possibility in the poetry world because Google-sculpting is not only a unique process (that while similar to earlier artistic processes, there is no real exact equivalent that I can think of) that is unique to the new medium, but: the act of Google-sculpting will likely become well known at least in the poetry world, and then will become a possible map or model for the writing process in general, or for the psyche/ imagination/ etc. in general, and will therefore cause unforeseen changes in the writing process/ poetic psyche/ imagination etc. in general.[23]

Again, this is a utopian reading of what technology can do for us and to us (co-evolution and the ‘advancement’ implicit in the term ‘evolve’) — not a nuanced reading of what technology does for/ to us (such as the implications of a virtual realm that’s not a mediator of reality so much as it is a mediated reality in toto — and what happens when this un-visceral reality is transposed to non-virtual reality with the capacity to make temporal and geographic distances negligible (e.g. war as video game / video game as war training)), nor is it concerned with how the technology works or who benefits financially from the use of it. Thomas Basbøll (who wrote an interesting essay in Fascicle 1 on Mohammad’s flarf technique by analyzing it in relation to the choices the poet could make but doesn’t) echoes this sentiment of Google as pathway to poetic enlightenment on the 12.02.04 entry of his blog The Pangrammaticon, comparing it to meditative breathing techniques & Pound’s aesthetics of distinction: ‘The sense in which both of the arguments pass from the unteachable mystery of poetry to the teachable craft, is exactly that which at some point brings us up to the Flarfomatic algorithms built into Google. You can quite distinctly program a machine to identify words and phrases and use this to establish a body-looseness (if you will).’[24] This is flarf as catalyst for a poetic state. I won’t argue with that. What I mean to emphasize is that it is also a conduit of corporate ideology. You might as well be using Jay Leno’s monologue to induce a meditative fugue — and what kind of sensibility is he transferring to the subconscious of those who fall asleep and slumber to his words every weeknight?[25] But even with such ideological prejudices (in this case my own) set aside, any ‘utilitarian’ tool is necessarily ideological: a good example is Slavoj Zizek’s use of toilet types as an excremental correlative counterpoint to Levi-Strauss’ semiotic analysis of food preparation (raw, baked, and boiled as not just methods but meanings):

In a traditional German toilet, the hole in which shit disappears after we flush water is way in front, so that shit is first laid out for us to sniff at and inspect for traces of illness. In the typical French toilet, on the contrary, the hole is in the back: shit is supposed to disappear as soon as possible. Finally, the American toilet presents a kind of synthesis, a mediation between these two opposed poles — the toilet basin is full of water, so that the shit floats in it, visible, but not to be inspected. It is clear that none of these versions can be accounted for in purely utilitarian terms: a certain ideological perception of how the subject should relate to the unpleasant excrement which comes from within our body is clearly discernible in it.[26]

The problem with Tost’s and Basbøll’s enthusiasm for the techno/ bio synthesis in poetry is that it recognizes Google as a utilitarian tool without also acknowledging its ideological architecture: it’s all ‘it’s the shit’ and no ‘what a piece of shit’: i.e. an excitement not tempered with apprehension,[27] a curiosity that isn’t also curious about itself, and (in Tost’s case) a declaration unsuspicious of its dismissal of a pre-existing process (e.g. Dadaist hat poems & collage) in favor of a clique-specific manifestation of it (Google-sculpting), which is then followed by a retroactive appropriation of that process (‘[flarf] will become a possible map or model for the writing process in general’). What we’ve got then is in fact a technical and internal co-devolution (in relation to their poetic predecessors) with the point-&-click technics of Google-sculpting acting as an externalization of a life of the mind that needs to get out more. This same devolution applies to all poems written according to chance operations determined by publicly available computer programs (e.g. online translators like Babelfish as well as word processor & email spellcheckers) that don’t also engage critically in the use of those programs. Here’s as discursive an analysis of the poetics of using Babelfish (already played out as a novelty back in 2003) that I’ve been able to find[28] from its web-savvy commentators: ‘You remember Babelfish poetry: take English text, use Babelfish to translate it, then translate the translated text back into English. Laugh. Repeat.’[29]

If there’s a difference between flarf and its progenitors it’s that Cage and Oulipo researched or created their generators of deterministic randomness, whether it be the I Ching, the weather, or mathematical formulas. They were aware of how each generator distinguished itself as a context and control variable, and their selection of each context and control variable was part of the content. Brian Howe (again, a poet whose work I admire) outlined the aesthetic justification behind his F7 series (which is generated out of and named after the spell-check function in Microsoft Word) on the 07.25.05 entry of his blog Slatherpuss: ‘F7 is akin to the works of the composer John Cage in numerous striking ways. For one, serial/ chance operations are utilized to recede the ego, imagination, and experience of the creator, thus freeing words/ sounds of an imposed value system and allowing for a more intense, less mediated experience.’[30] Which sounds great — except his F7 series is mediated by the imposed value system of Microsoft engineers. Recession of the ego and incorporation of chance into aesthetics are certainly Cagean — but so’s the undermining of ideology, and the prejudices and self-justifications it obscures.[31] Howe is aligning himself with Microsoft without exploring the problematics of that alignment: not just what does it produce, but how, and why? A collage is always the creation of something totally new — but it’s also an engagement with what the pieces of it once were, as well as with the tension between what each piece was in relation to what each piece is, and the tension between the pieces, what they were and what they are, in and out of context, and the tension between the contexts. In other words, aside from an immediate visceral appreciation, what’s interesting about collage is the relationship between the material as originally contextualized and the material de- and re-contextualized — all of which continually shift as the collage is made and experienced by an audience. It’s an engagement with intersecting lines of reciprocal complicity: how do the lines compromise, enhance, ruin, challenge, obscure, rewrite, and ignore each other — in tandem, counterpoint, and isolation, alternately and all at once?

The flarfists may be aware of the webpage from which they borrow material, but the only reason they’re aware of that webpage is because Google (or AskJeeves, or Yahoo!, or...) showed it to them — so the question is, are they aware of why they’re aware of that webpage? Do they wonder how it is that their poem is determined as it is — that is, of the process at work on their work by an outside force, one not divine or natural but corporate? This is a fundamental aesthetic concern as well as a socioeconomic one. And a process that draws so much attention to itself as process (since the end result is often intentionally discordant with overt tonal & syntactical juxtapositions) undermines its credibility if it propagates a lackadaisical attitude toward its own mechanics.

A look at Tost’s article[32] in Fascicle 1 on K. Silem Mohammad’s use of flarf substantiates this lack, beginning with his understanding of flarf’s capacity to re-contextualize: ‘But I will argue that the re-imagining of source, and the reader’s knowledge of the source of Mohammad’s language, is perhaps the great realization of these poems.’ A re-imagining of the source as the webpage itself, sure — but a re-imagining of the search engine, which, for Mohammad, is the primary source? Neither Mohammad as poet[33] nor Tost as critic address the socioaesthetic implications of his process: the search engine, not the webpage, is the source of Mohammad’s poems. It mediates and presorts the information for him, already juxtaposed and abbreviated, which he then edits. This editing may be a ‘re-imagining’ of the language, but Google beat him to the punch and did it first. The most significant re-imagining of Mohammad’s primary source that’s occurring is an inadvertent classification of the search engine as ‘silent’ in a Cagean sense. Tost confirms this silence by quoting Jack Spicer in support of flarf’s enabling of the Other: ‘I think the source is unimportant.’ Which is both true and not true: true because anything can be a catalyst for poetic engagement with the Other; false because the source affects how that poetry takes shape, which means that Google determines the poem that happens as much as the poet does.

Tost’s limiting interpretation of Spicer’s comment also shadows his claim that flarf ‘could be presented as a parody of Pound’s great Modernist innovation where, in Hugh Kenner’s words, his vortex is “not the water but a patterned energy made visible by the water.” Mohammad’s poems and lines aren’t the search words but are patterns made visible by the search words.’ But what patterns are these, and just how visible are they? Tost is talking about lexical, morphological and syntactical patterns and Mohammad’s method of assembling them — what he’s not talking about are the invisible a priori patterns of which these patterns are a product: the selective algorithms of search engines, the manipulation of these algorithms by search engine editors, and the exploitation of them by corporate web designers. This blindness is ideological, made explicit when Tost falls back on the idea of the Internet as a manifestation of democracy in its pure form: ‘For the contemporary poet’s imagination, technology is a democratizing force; a poet is pushed to acknowledge the divide between his or her poetic presentation of everyday speech and the actuality of everyday vernacular as it occurs in chat-rooms, personal websites and the like.’[34]

But couldn’t Mohammad’s flarf be considered a capitalist, as opposed to democratic (though the two are often conflated), co-opting of ‘non-poets’ for poetic gain, as if they didn’t have a voice before he gave them one by taking theirs? That Elvis invented rock ‘n’ roll? And isn’t Mohammad guilty of enacting his own criticism (quoted by Tost) of ‘the effete peripatetic poet safely above a scenic view of the countryside and its filthy horizon’ by surveying his materials from the Godlike perspective of Google?[35] He’s not in the chat rooms or forums, he’s not engaging in the stutters and misspellings and expletives and non sequiturs in real-time — he’s not participating in the language as a living thing or process, but as an artifact, and with a lexical condescension as evidenced by his choice of the pejorative ‘drivel’ to describe chat room semantics when relating flarf’s genesis.[36] He’s ‘accessed’ nothing, as Tost claims; what the search engine does is survey, i.e. take ‘a scenic view’, so that the language has already been distanced and re-contextualized by the time it reaches him. What’s perhaps worse than an ‘effete peripatetic poet’ is an effete sedentary poet making claims that he’s getting his manly hands dirty by resting them on a keyboard and clicking ‘search’. In this case the ‘filthy horizon’ is his tacit belief that he’s participating in the language when his only direct engagement is with Google’s indexing of that language.

Tost also compares flarf to Benjamin Friedlander’s pop-appropriated ridiculing of the ideology of the Buffalo Poetics Listserve from within, presenting it as a radical new aesthetic shaking up poetics and the poetry community: ‘Like Flarf, Friedlander’s project addresses and attempts to counter a certain insularity and preciousness in the poetic community.’ But my argument is that flarf, as it’s been practiced and criticized, is in fact a product and a perpetuation of that same insularity. It’s as ineffectual as the Mary Oliver poem that Tost derides for its ‘assumption of an unusual degree of safety and comfort’ as well as its ‘assumption of ample time for both leisurely experiencing nature and contemplating its importance.’ What could be more safe and comforting than the interface of a computer screen to those who have ‘ample time for both leisurely experiencing’ the Internet ‘and contemplating its importance’? The old school classist distinction Tost is positing is actually a demographic one: to anyone with even a peripheral relationship to consumer culture, the Internet is as common as the trees. A similar case of the pot calling the kettle black is Mohammad’s observation (from the 09.17.05 entry on {limetree}) that

plenty of “New Formalist” poets are politically aware in their personal lives, even as their work is complicit with a retrograde literary ideology — an ideology that nostalgically fetishizes the most superficial aspects of form, at the expense of appreciating fully the contexts within which such forms were put to use in their own historical moments.[37]

Which, once again, is more or less a diagnosis of what’s wrong with flarf as Mohammad and others create and critique it. Tost ends his analysis by assigning Google ‘a role somewhere between Personism’s telephone and Projective Verse’s typewriter: a new instrument for both gathering information and for re-imagining the construction of a poem.’ Certainly this is true. But what kind of instrument? And how does it gather information? And in what ways (and why those ways) is it re-imagining the construction of a poem?

So let’s give Google the once-over already.[38] My reason for doing this is not to discredit it as an effective generator of poems, but to call attention to the problematics of using it as such.

How Google Works

Google has not publicly released its algorithms or given a specific breakdown as to how it weighs the individual factors that determine the relevancy of a webpage in relation to the search term. But there’s a burgeoning industry of companies whose job it is to reverse engineer the process and figure out the mechanics of Google so as to exploit it; the idea being to optimize website structure, design, and copy according to the most commonly typed linguistic forms of a given keyword phrase in order to increase the website’s page ranking for those keywords, which in turn increases its exposure, hits, and profit. This is referred to as Search Engine Optimization (SEO)[39], and the general question these companies (and the flarfists should) ask is the following:

How does Google decide on the relevancy of a page to a given keyword? i.e. How does it rank what it thinks you’re looking for?

Google sends out ‘crawlers’ (aka ‘spiders’) across the web to find pages in which those keywords are used. Then it calculates relevancy with its PageRank algorithm based on a handful of factors, including:

Popularity: Number of hits a website receives per day.

Interconnectivity: Number of links to a website from other websites, as well as from pages within the site (the text of the link is called ‘anchor text’ and is included in calculations of keyword density for the linked-to page). Google interprets a link from page A to page B as a vote by page A for page B. But not all links are equal: those from ‘relevant’ web pages, of certain domains (.gov, .edu., .mil), and those with a high page rank are weighted more heavily than others.

Potency (aka ‘keyword density’): keyword appearances in anchor text, URL, title tag, meta tags, alt text, and body of webpage; font size, typeface, and area on the page in which the keywords appear; and the number of times they appear in relation to total text on the page. The title tag tends to be the most heavily weighed since many search engines use the keywords found in the title tag to index the web page. But Google’s spiders (collectively called Googlebot) can’t read text contained in images or pop-ups (and most search engines have trouble with Flash and JavaScript, and can’t even read pages with dynamic URLs), so these aren’t included among its calculations of keyword density. Which means its algorithms aren’t equipped to assess the full textual content of webpages and thus determine ‘relevancy’ in any kind of comprehensive sense, even if the code were somehow ideologically transparent.[40]

[It should be noted that because of their archived and heavily intra- and interlinked sites, the blogs of the virtual poets have an unnaturally high keyword density and interconnectivity built into their design; Google favors their daily compounding of intralinks and keywords without compensatory handicapping.][41]

When determining relevancy, PageRank filters the results, for the following reasons:

To main integrity of process: Google will penalize pages that attempt to manipulate search engine spiders with an unnaturally high keyword density (this varies between search engines, but anything over 5% is suspect) or keywords that are the same color as the background and therefore invisible to the browser. These techniques and others are known as ‘spamdexing’, and Google buries their ranking accordingly. But as long as companies follow Google’s ethical guidelines, manipulation of the system is not only allowed but encouraged and rewarded with a high page ranking.[42] If you search for a keyword that’s in any way commerce related, chances are that much of what appears on the first page has been optimized in some way to attain that position — and what’s not commerce related? And if you take the time to click through the Google results to the 100th or 200th link and beyond, you’ll find many of the same big money sites indexed more than once for multiple pages, or the same article hosted across multiple sites, which makes for a rather homogenized random generator that has trouble distinguishing between relevancy and redundancy. But optimized or not, money matters to Google: a search for ‘poetry’ pulled up 20 results on the first page, 10 of which were sponsored links (with a link to more sponsored links), with the 10 ‘natural’ links being bookended by infamous vanity scammers poetry.com[43] and multimillionaires Poetry magazine.

To enforce a moral code: Google’s SafeSearch blocks web pages containing explicit sexual content from appearing in search results. Its default setting is ‘moderate filtering’. To search with no censorship filters the user must go into Google preferences and manually change the option. But even this won’t bring all of the sites with sexually explicit content to the foreground, since many of them fall under Google’s definition of ‘spam’ and remain exiled to the far reaches of search engine irrelevancy.

Then, after calculating relevancy and filtering the results, Google presents the results to the user. If the keyword is considered marketable, ‘sponsored links’ will appear above and to the right of the search results. These sponsored link spaces are allotted for certain keywords via Google’s targeted advertising program called AdWords, and go to the highest bidder: companies bid on a pay-per-click number, which means they only pay Google when someone clicks on the link as a result of a keyword search. The highest bidder gets the highest placement, which is congruent with Google’s hierarchical thinking (since vertical placement is also a factor in calculating keyword density). The obvious implication is that the Internet is evolving from a facilitator of information into a generator of money: search engines are selling keywords, and websites are structuring themselves accordingly. This isn’t democratic — except in the most cynical, contemporary sense of the term, in which free elections are routinely won by the candidate with the biggest ‘war chest’.

Regardless, I suspect most search engine collagists will consider ‘sponsored links’ invalid and disregard them — but the ‘natural’ search results are still contextualized by these paid results, and the paid results are as much a part of the zeitgeist since ‘consumers’ respond to advertisements, in and of themselves, as a validation of credibility, based on the logic that if a company has money to advertise then it’s legit, and might also offer lower prices in an effort to undercut those who can’t afford to do so.[44] To ignore sponsored links when search engine collaging is a decision to censor this part of the source material and culture at large. Which is fine — but to be unaware of this is to be unaware of one’s ideological prejudices and marginalizations. It’s to have learned nothing from Cage’s observation that there is no silence: ‘silence’ is just the name we give to the sounds we ignore.

How to Proceed

With all this in mind, here are some issues, in no particular order, that any practitioner of search engine collage should consider before beginning:

To use Google to sculpt poems and/ or promote the writing of poems in this way is to be an unsolicited advertiser for a multi-billion dollar company. Not that this is in and of itself an anti-poetic or anti-human act. I am not disputing the relative benefits and detriments of Google in relation to other corporate entities. Nor is it disputing the decontextualized aesthetic merits or visceral impact of the poems themselves. I’m just calling attention to the fact that the makers of the Google search engine have a vested financial interest in the popularity of their program. And to use corporate algorithms, and especially to use and promote a single corporate algorithm by name (e.g. the phrase ‘Google-sculpting’), is to be complicit in the corporate agenda of propagating the use of their product — in this case letting it permeate the arts (& all spheres of human life) without analyzing the pros & cons of doing so. Which means the issue here is not so much the use of Google as it is the passive acceptance and uncritical implementation of it.

To use and/ or promote a search engine as a breakthrough generator of deterministic randomness is to conflate process with an arbitrary form of that process. A primary artistic rationale and impulse behind search engine collage is the desire to engage the Other outside & thereby inside oneself. But Google-sculpting as an individual project or culturopoetic phenomenon necessarily atrophies that impulse; instead of maintaining its initial aura as an endless generator of random cultural code (which is itself inherently dubious), the perpetual use of it turns the Other into a Familiar.[45] A systematic analysis of results for one keyword across multiple search engines might be interesting as a lexical variation on Steinian syntax — but how many times can you do this without becoming totally bored with yourself? And how systematic is it when you don’t investigate the criteria and resource pool of those search engines, and how interesting is it when you don’t consider the results in relation to that criteria and resource pool?

To use and/ or promote a search engine without question is an implicit acceptance of it as an arbiter of relevance. This is like compiling a list of what’s going on in the world from only the U.S. network TV news shows, without acknowledging the biases inherent in their selection processes. Google is not the zeitgeist, nor is it an indifferent and all-inclusive database of it. Google, as a generator, is a corporate algorithm that ranks webpage relevancy (from a limited cross-section — i.e. what it bothers to index — of an already limited segment of the zeitgeist — i.e. Internet users and web content) based on its own idiosyncratic definition of ‘relevancy’ and, in tandem with corporate web designers, manipulation of the results.

Don’t Be Evil

And so for collagists, it’s important to recognize the supposed silences in and around the material or else perpetuate a kind of redundancy in the act of creation. These silences are the answers to general questions we should ask, before/ during/ after, so as to know just what it is we’re recontextualizing in the act of search engine collage. But even before questioning Google’s selection and arrangement of information, there’s the question(s) of the Internet itself:

What is the ratio between the various types of websites (commercial, academic, government, blog, & personal) on the Internet? How many of each are indexed by search engines? Which websites are linked to each other? Who owns these websites? How many of them have been optimized? How many people use the Internet? How many use search engines? How many use Google? What demographic segments do they represent?

Etc. and so on. Not that these questions can necessarily be answered; but to not address them is to place an a priori trust in the agenda and integrity of search engine companies, which is not only an aesthetic fallacy but also a compromise of one’s privacy. Google, in particular, has marketed itself as a benign facilitator — its corporate motto is ‘Don’t be evil’[46] — and its zealous enthusiasts put their money where their mouths are when Google became a publicly traded company in 2004.[47] Through corporate acquisitions and its new investment muscle, Google is now set on indexing not just the information available on the Internet, but in universities, libraries, and email correspondence — in a sense, making a virtual Borgesian map of the entire world. From a utopian point of view, free information is great. But the situation is necessarily complicated, as evidenced by the outrage from authors and publishers over the copyright infringement of Google’s Library Project, as well as from consumer advocates who protest its scanning of personal email content (via Gmail), which content it stores and indexes for purposes of targeted advertising. It also stores and indexes what keywords you search for, and when you search for them, as well as your geolocation, via your computer’s IP address and the cookies it places on your machine (which store your preferences and track patterns of how you search — and don’t expire until 2038).[48] If you sign up for Gmail it connects your name to what you search for via your IP address.[49] Which means that any searching you do in the virtual realm leaves a digital, catalogued imprint directly traceable to your name, and is available for whatever authority finds it valuable. What reason does Google have to index this information in such personal detail, and for as long as it does? Does it specifically stipulate that it won’t sell it to companies for profit or trade it to governments for political leverage?[50]

Google’s official privacy policy states ‘We do not rent or sell your personally identifying information,’ immediately qualifying this statement with ‘We may share such information in any of the following limited circumstances,’ one of which is ‘[If] We conclude that we are required by law or have a good faith belief that access, preservation or disclosure of such information is reasonably necessary to protect the rights, property or safety of Google, its users or the public.’[51] In other words, the information is confidential, but not really. The obvious issue here is that terms like ‘good faith’ and ‘necessary’ and ‘protect the rights’ and ‘safety’ and ‘public’ are contingent on the ideology of the speakers in question. How do they define these terms? How do they justify their actions with their definitions? Notice the syntactical priority given to itself in the sequence ‘to protect the rights, property or safety of Google, its users or the public’, as if its primary concern is to save its own ass; or the specificity of ‘we do not’ in regard to selling personally identifying information, as opposed to ‘we will not’. This evasive phrasing is not a promise of future action but a description of present action, and is analogous to responding to the question ‘Do you steal from people?’ with ‘I am not stealing right now.’ And what promise do we have that even this dubious policy will remain in effect, in the event of a corporate merger or buyout? ‘In the event of a transfer of ownership of Google Inc., such as acquisition by or merger with another company, we will provide notice before any personally identifying information is transferred and becomes subject to a different privacy policy.’[52] This in no way suggests there’s a way to prevent a transfer of your personally identifying information. Nor will they let you see what they’ve collected about you.[53] What Google is promising is to let you know when it’s giving it away.

All of this is not meant to vilify Google per se, but to call attention to the problematics inherent in its claim of ‘Don’t be evil’ as a guiding principle, as well as in the Google-sculptors’ implicit acceptance of that claim, and complicity with it, by using Google as a poetic generator without also calling that use into question. If it were just a concept in a theoretical vacuum Google might be inclusive in scope; but its sorting principle is exclusionary and artificially hierarchical, manipulated as it is by its creators and the people who create its content in a joint effort to facilitate marketing (in addition to being limited by its technological deficiencies). Thomas Basbøll is aware of this even if he isn’t aware that he’s aware. From the 12.04.04 entry on The Pangrammaticon: ‘If marketeers are putting words together at random, then perhaps Google will not serve the purpose I had hoped in regard to prosody. I thought Google could determine interesting proportions in various word combinations, distinguishing orthodox from original grammar (in order, of course, to avoid the latter). I do not want to give the power to determine orthodoxy to e-bay’s henchmen.’[54] Except he’s already giving the power to determine orthodoxy to Google’s henchmen. This is the problem inherent in flarf that none of its practitioners or published critiques have recognized — it is already ruined. Not that it wouldn’t be interesting to explore the ruins. But as it is, flarfists might as well be riding Disney’s ‘It’s a Small World After All’ theme park ride and equating themselves with Magellan. Basbøll senses this as the future instead of the present:

Can self-respecting poets really apply the luminous ideoplastic of “the basilica nearby the motorcycle oil” or the “frippery at some saggy panorama” to their work knowing what their source may be? Are we back to searching through the whole language within, word by word, hoping to be blessed by the muses?[55]

The first question is unintentionally self-reflexive, but it also highlights the real issue here: it’s about awareness, not power. There is always a power determining orthodoxy in any ‘random’ generator — but there’s also a power determining orthodoxy in any method of communication. This is at the heart of the 20th Century avant-garde from Duchamp to the L=A=N=G=U=A=G=E project. The aim is to elude orthodoxy by calling attention to it, disassembling it, and constituting something new from the pieces.[56] Basbøll’s fatalistic solution is to take his ball and go home, ‘hoping to be blessed by the muses’ — but even muses are infected with orthodoxy. Even thinking is a form of communication, with oneself. There is no escape. This is our given. It’s our starting point, as artists and human beings; and yet it’s out of such points of total futility that true revolution is possible, in which, to paraphrase Breton, we conceive the impossible (i.e. radical change) and believe it in spite of ourselves. One then acts, and evokes, accordingly. In the meantime what’s both prudent and necessary is to proceed toward/ along a goal/ path of awareness of the paradigmatic structures within which we live; not to ignore or simply critique but to engage; otherwise, we risk believing, via a cavalier and unknowing complicity, or else a distanced finger-pointing, what the architects of these structures want us to believe: that those structures are the natural way of things and free of ideological intent, or that we ourselves are free of those structures and ideologies. But with a guiding principle of awareness as opposed to randomness or democratic or experimental or egolessness or any of these other incomplete in themselves buzzwords the flarflists have evoked, search engine collage can be an effective poethical[57] project. And one that somebody somewhere should be doing, right now, as Google expands and insinuates itself further and further into our lives. What’s needed is less avant-garde, post- or otherwise, and more of what poet/ critic Joan Retallack calls Cage’s avant-pragmatism.[58] What is Google really telling us? And what are we telling each other when we use it?