Code Poetry in Motion: E.E. Cummings and his Digital Grasshopper

John Freeman (bio)

Abstract

This essay argues E. E. Cummings’s “r-p-o-p-h-e-s-s-a-g-r” (1935) anticipates the contemporary practice of experimental writing known as codework. Encoding through typographical means the action sequence of the grasshopper’s leap, Cummings transformed his mechanical typewriter into the equivalent of a hardware device supplied with the necessary software for running the poem as a program. The three permutations of “grasshopper”—”r-p-o-p-h-e-s-s-a-g-r,” “PPEGORHRASS,” and “gRrEaPsPhOs”—reflect a block cipher method of encryption, each one comprising a subkey governing the operations of its surrounding textual arrangement. Like the punch cards driving Jacquard’s loom and Babbage’s Analytical Engine, these subkeys weave for us a digital grasshopper.

We may say most aptly, that the AnalyticalEngine weaves algebraical patterns just as the Jacquard-loom weaves flowers and leaves.—Ada Lovelace

Strangely enough, the poet who once asked, “How numb can an unworld get?” and answered, “number,” also bequeathed to us the world’s first fully operational digital poem. Enacting its subject (the grasshopper), Cummings’s poem “r-p-o-p-h-e-s-s-a-g-r”¹ makes a leap from the analog to the digital (and back again). The possibility of such a leap is only partly entertained by Rita Raley in her essay “Interferences: [Net.Writing] and the Practice of Codework.” She does so in discussing codework, a mingling of natural language and pseudocode that creates “a text-object or text-event that emphasizes its own programming, mechanism, and materiality.” Turning her attention to Cummings and other typewriter experimentalists of his time, she asks us to picture them “upgrading their medium and exchanging their typewriter keys for the units of programming languages.” The result, she claims, “would in part resemble the contemporary mode of experimental writing and net.art called ‘codework.'” Prematurely abandoning her comparison, however, she maintains that the difference between “e. e. cummings’s ‘r-p-o-p-h-e-s-s-a-g-r’ and a net.wurked text … is the difference between the typewriter and the computer, the difference of what the medium allows.” On further inspection, however, Cummings’s poem, digitally expressed, anticipates the operations of the code work in ways that might very well cause Raley to acknowledge its status as a code poem.

While it is true that the lack of an appropriate medium imposes definite constraints on technological as well as aesthetic innovation, we have only to consider Charles Babbage’s proposed Analytical Engine in combination with Joseph Marie Jacquard’s punch-card-automated loom to verify that such constraints can be overcome while working with the materials and concepts at hand. A refinement of his Difference Engine (a design for a mechanical computer), the Analytical Engine was conceived as a general purpose computer capable of conditional branching (executing different instruction sequences in response to new data) and equipped with memory. Fascinated by the Analytical Engine, Ada Lovelace, the daughter of poet Lord Byron, worked closely with Babbage on the project. In fact, she was commissioned by him to translate a speech he had given to Luigi Menabrea, the future prime minister of Italy, who had been so impressed with Babbage’s presentation he had transcribed it. Lovelace’s “Sketch of the Analytical Engine Invented by Charles Babbage” (1843) far exceeded expectations for either a sketch or a translation, as she contributed a great deal of her own analysis and insight to the manuscript. Aware of the punch cards that Jacquard employed to govern the operations of his looms, she proposed employing them to direct the operations of the Engine, thus expanding its capabilities from solving mathematical problems to performing other feats of “composition,” as set forth in the following prescient speculation: “the engine might compose elaborate and scientific pieces of music of any degree of complexity or extent” (Lovelace 694). She foresaw the concept of Band-in-a-Box a century-and-a-half before its time.

With her penchant for “poetical science” (Chiaverini 426), Lovelace is portrayed by James Essinger as a boundary-breaking thinker who operated in a “thought … domain of the intellectual prehistory of the computer” (Ada’s Algorithm 172). Indeed, her notion of using Jacquard’s punch cards to conduct the mathematical operations of Babbage’s engine shows that existing technologies and mechanical operations can help define and be put to the service of envisioned ones. (Lacking the necessary funding, the Analytical Engine was never built in Babbage’s lifetime. In ten years, however, researchers in England hope to build a working model based on his sketches.) In discussing the handling and storing of these punch cards, Essinger signals their connection to modern practices: “Weavers would keep these chains [of cards] in a storeroom whose function was very much the same as that of the library—or we might say software library—which Babbage was proposing to create” (Jacquard’s Web 91). Citing a passage from Lovelace’s description of the Analytical Engine, in which she clearly anticipates “the fundamental difference” between computer software and data, Essinger observes: “This passage could almost be an extract from a modern computer manual, but written by a Victorian” (Jacquard’s Web 122). In similar fashion, “r-p-” (1935) anticipates digital poems, demonstrating that “what’s past is prologue” to the future. In setting out to encode through typographical means the action sequence of the grasshopper’s leap, Cummings transformed his mechanical typewriter into the equivalent of a hardware device supplied with the necessary software for running the poem as a program.

1. Punc[h]tuating the Analog/Digital Interface

At first blush, Raley’s definition of codework would seem to mark an unbridgeable divide between Cummings’s production and those of the codeworkers, whom she describes as employing the idiolect of the computer to create “the contemporary mode of experimental writing and net.art called ‘codework.'” Nonetheless, Cummings, like Lovelace and Babbage, practiced his own brand of “poetical science,” fashioning, bricoleur-style, his own idiolect, and serendipitously anticipating the coding functions of the computer through his typewriter experiments. For example, the code poet Mez expressly strives “2 spout punctu[rez]ationz reappropri.[s]ated in2 sentence schematics.” We find similar “puncturings,” “reappropriations,” and twisted, entangled syntactical arrangements in this poem and elsewhere in Cummings’s œuvre (as in these opening lines, performatively reenacting this time the onset of a cat’s leap: “(im)c-a-t(mo/b,i;l:e”). As Aaron Moe observes, “the word ‘c-a-t fall[s] into the word immobile, thereby breaking the immobility in half in order to declare I’m mobile” (114). In similar fashion, the poet sets his grasshopper in motion.

Cummings engages in what Raley describes as the “art of the code, in which the code used to produce the work seems to infiltrate the surface, the former domain only of natural languages and numeric elements.” Here, Cummings uses “calculated dislocation”—misplaced, displaced, disorienting grammatical markers—to indicate the encoding, governing elements on the surface of the text (Friedman 109). Breaking up words into smaller units, he subverts orthography in favor of an aesthetics of “cacography.” Like Mez, he aims to “disturb, disorient, and defamiliarize, to shift ‘the units of information and communication from the usual and expected to the cryptic'” (Raley). As Geoffrey Leech informs us, constructions that would be dismissed by linguists as “unmotivated deviation—a linguistic ‘mistake'”—are foregrounded here. Norms of language become a background while “features which are prominent because of their abnormality are placed in focus” (30). Leech points out how this process of defamiliarization results in “impeding normal processing” (4). Impeded as well, readers must work out for themselves what new decoding procedures must be followed in interpreting the poem.

There is a somewhat prescient correlation between Cummings’s use of punctuation and its current functions in computer coding. Perhaps this is not so prescient after all, since the typewriter experimentalist and the programmer work from keyboards with largely similar arrangements of letters, numbers, and symbols. As programming developed, it could take advantage of punctuation marks, which, as Roi Tartakovsky observes, possess a fixed institutional character as defined by handbooks while also being “amenable to appropriation, to exploitation, and to projection.” Their quality of being “void of semantic content” means they can be repurposed. The unconventional punctuation frequently used by Cummings causes a similar recoding to occur. Indeed, by dislocating punctuation marks from their traditional moorings, he casts them adrift, opening a space for new usages. Infiltrating the domain of natural language, this redeployment of grammatical signs bears comparison to script prompts (for example, quotation marks to wrap the string, parentheses to mark functions, the period to designate a file name (.exe), and the semicolon to indicate the end of a statement [as in “.grasshopper;”]). To riff on Mez, such retronyms “punc[h]tuate” through natural language, infiltrating it with new meanings, new encodings. Re-programming language, Cummings fortuitously at times lights upon usages yet to be determined, a brand of retrofuturism in the making.

This infiltration marks “r-p-” as a prime example of the reverse remediation process that N. Katherine Hayles defines as “the simulation of medium-specific effects in another medium” (73). Indeed, the poem emulates many of the qualities she specifies as characteristic of electronic hypertexts. To the extent that it resists the traditional coding procedures by which we read a text, it evidences a supra-analogical dimension, what Hayles describes as “deeper coding levels” beneath “the flat surface of the page” (76). Raley’s description of Mez’s use of brackets and periods to split words “into multiple components” could equally be applied to Cummings’s own stylistic practices. This feature aligns Cummings’s work with one of Hayles’s categories for the electronic hypertext: “Generated through Fragmentation and Recombination.” Situated in the analogical/digital divide, Cummings’s poem favors the digital register, in which “the fragmentation is deeper, more pervasive, and more extreme than with the alphanumeric characters of print.” Where the fragmentation of the hypertext “takes place on levels inaccessible to most users,” however, a poem like “r-p-” foregrounds it (Hayles 77). Readers not only perceive the fragmentation and recombination at play, they must also negotiate these features, qualifying “r-p-” for another of Hayles’s electronic hypertextual features: navigability. The poet’s experimentation with linguistic codes anticipates code poets’ own disruptive procedures and contributes to the defamiliarization effect of reading Cummings’s poem. Readers are forced to decode the poem’s non-analogical elements, to locate and make sense of the unconventional program running its operations. The reverse remediation accomplished here anticipates the “Cyborg Reading Practices” that Hayles will outline more than a half-century later (74).

Upon closer inspection, “r-p-” surpasses Mez’s codework and other inoperable forms of code poetry by satisfying the criterion that code poetry purists insist upon: that such works be executable. In “The Poetry of Executable Code,” Roopika Risam distinguishes “between code that is operational and has depth and code that is isolated on the surface of a text.” As we shall see, “r-p-” displays the operationality and “depth” qualities that Risam requires of codeworks. Raley’s discussion of code poet Graham Harwood’s transcription of William Blake’s “London” (1791) is instructive in illustrating more clearly the distinction between operational and surface code poetry. In the original, the speaker walks the city streets, lamenting:

In every cry of every Man,
In every Infant's cry of fear,
In every voice, in every ban,
The mind-forg'd manacles I hear.

Harwood’s poem “London.pl” is titled as a file name and employs Perl code. Its rendering is described by Raley:

Aside from the comments, it contains a definition 
of what in Perl is called an "anonymous array," i.e. 
a variable storing several values at once, called 
"@SocialClass," a database (or, in programmer's 
lingo: "nested hashtable") "%DeadChildrenIndex",
and two sub-programs ("subroutines") 
"CryOfEveryMan" and "Get_VitalLungCapacity".

Noting that Harwood translates Blake’s poem into “symbolic machinery,” Raley cites the following passage, written in source code, as a “macabre” reprogramming process of the original:

# Find and calculate the gross lung-capacity of the children

# screaming from 1792 to the present

# calculate the air displacement needed to represent the public

# scream

# set PublicAddressSystem instance and transmit the output.

# to do this we approximate that there are 7452520 or so faces

# that live in the charter'd streets of London.

# Found near where the charter'd Thames does flow.

Drawing upon imagined databases and numerical “tabulations” of despair, the command prompts drive home the narrator’s distraught observations. Although it incorporates snippets of Blake’s poem (“charter’d streets” and “charter’d Thames does flow”), Harwood’s program functions more as an outside commentary or supplement to the original. It represents a brand of codework Raley describes as one “that incorporates static, non-functional elements of code into the surface, or ‘Interface text.'” While suggesting imagined databanks the poem might draw upon and sub-programs it might run, it does not offer—text-editor style—any sense of how the original poem was encoded or any indication of the program that governs its operations.

2. “r-p-o-p-h-e-s-s-a-g-r” by the Numbers

Exploring how “r-p-” is coded and transmitted—or digitized, a computational approach—reveals a good deal about the poem as an operating system. If we imagine for a moment the poem’s appearance on a computer screen, we can apply Espen Aarseth’s concept of the scripton to the surface level image we encounter. His coinage “texton” references the underlying code that reflects and governs its operations. As Hayles informs us, “In a digital computer, texton can refer to voltages, strings of binary code, or programming code, depending on who the ‘reader’ is taken to be” (81). Assigning numerical values to the word “grasshopper” can help us to chart its digital permutations in the course of the poem and will reveal its encoded functions as we approach the binary level.² The letters in the word “grasshopper” can be represented in the following numerical fashion (see fig. 1):

Fig 1. The word “grasshopper” realized numerically.

Cummings’s poem operates like an analog-to-digital converter, employing discrete or continuous values to represent information, most notably by “sampling” single letters or blocks of letters from its subject’s name, encrypting them, and then “transmitting” them, as represented by the three scrambled forms expressing that name: r-p-o-p-h-e-s-s-a-g-r, PPEGORHRASS, and gRrEaPsPhOs.

Aligning “grasshopper” with “r-p-o-p-h-e-s-s-a-g-r,” we can arrive at a digital conversion of the poem’s first word (see fig. 2):

Fig 2. Digital conversion of “r-p-o-p-h-e-s-s-a-g-r.”

My colleague and mathematician Jeffrey Boats solved the problem of which numbers to assign to the two “r’s” by pointing out that “grasshopper” contains double consonants and blends (e.g. pp, ss, and gr). The first “r” (following “g”) should therefore be assigned a “2” and the final letter “r” an “11.” He further pointed out that the three forms can be expressed by a table of permutations, as the rendering of “r-p-o-p-h-e-s-s-a-g-r” in the example above. Transposing the number values of the bottom column with the corresponding numbers above them (1 to 11, 2 to 8, and so forth) renders the scrambled term. Reversing the direction of transpositions (11 to 1, 8 to 2, and so forth) renders “grasshopper.”

________

While Michael Webster sees “a marked degree of symmetry” in the above permutation’s placement of an “r” at each end and an “e” in the center (“Prosody” 134), “r-p-o-p-h-e-s-s-a-g-r” poses the most difficulty of the three arrangements for readers in their initial encounter with the poem. Cummings’s obfuscation of natural language in scrambling the “r-p-o-p-h-e-s-s-a-g-r” signal has its modern-day equivalent in obfuscation coding, the deliberate use of source or machine code to confuse or lead astray those attempting to use it. For Nicholas Montfort, “All obfuscations … explore the play in programming, the free space that is available to programmers” (197). There is an aesthetic pleasure, Montfort maintains, in exploring “the gap between human meaning and program semantics” (193). Webster expresses a similar assessment: “These visual, linguistic, and mathematical calculations and playings on/with the poem inevitably result in an intense engagement with a living and moving poem” (“Prosody” 138). There is little doubt that the midsection of the poem affords readers a great deal of free play as they try to gather in and make sense of the scattered elements swarming around there.

In Montfort’s appreciation of this practice of willful deception, he writes that “it throws light on the nature of all source code, which is human-read and machine-interpreted, and can remind critics to look for different dimensions of meaning and multiple encodings in all sorts of programs” (198). In this regard, we can reverse-remediate Florian Cramer’s question: “Can notions of text which were developed without electronic texts in mind be applied to digital code, and how does literature come into play here?” We can now ask how the digital code of electronic texts might illuminate the operations of a [hyper]analogical text. Indeed, with “r-p-” we are dealing with what Raley might describe as “an algorithmic poem, which changes the system in a materially visible way.” Crossing over the analog/digital divide, Cummings deserves to be featured in an anthology of twenty-first century code poets.

The conversion of letters to numbers reveals the programmatic coding operations behind the three expressions, which constitute a digitized refrain for the poem. (These numbers could in turn be expressed by the ones and zeros of machine code, but since I want to communicate with humans and not machines, it is more expedient to keep them in this format.) The progression of the three forms leading up to “grasshopper” can be likened to an analog-to-digital and digital-to-analog transmission and reception process. Of course, since the “grasshopper” signal is not sequentially transmitted and reconstructed until the end of the poem, its analog elements have already been encoded as “r-p-o-p-h-e-s-s-a-g-r.” when readers first encounter the poem. As we shall see, this scrambled construction and its digital equivalent (11 8 7 9 6 10 4 5 3 1 2) represent an extremely low sampling rate.

Each manifestation of “grasshopper” constitutes a new stage in an analog-to-digital conversion process. The ensuing forms, “PPEGORHRASS” and “gRrEaPsPhOs,” are products of an increasingly accurate sampling rate. The appearance of “grasshopper” at the end of the poem signals that the digitizing process, having been reversed in the course of the poem, has reconstituted the encrypted analog “message” for reception by the readers.

As the analog-to-digital conversion progresses from the low-sample “r-p-o-p-h-e-s-s-a-g-r,” we find a more orderly arrangement in the next manifestation (see Fig. 4).

Fig 3. Table of Permutations for PPEGORHASS.

In the above permutation, there are formations consisting of three- and four-letter block constructions, PPE (8, 9, 10) and RASS (2, 3, 4, 5). These motion more to the final formulation: G RASS H O PPE R.

Fig 4. Formatting of PPEGORHASS showing block formations.

As a further sign of increasing orderliness, we can even see a path toward a more complete reconstruction by “cross-switching” the 1 and 11 within the inner part of the two blocks and moving the 7 and 6 to the outer part in a square-dance fashion (see fig. 4). A third operation, reversing the two columns and shifting to lower-case format, would result in “grasshopper.” In his handwritten draft, Cummings originally played with the notion of including either “andwro ngwayr ound” or “wrongwayroundfully” in this section of the poem. It must have occurred to him, however, that the letters (as demonstrated by the digitalization) were already doing the work of directing the operations of the poem for him. Expressing what Wendy Hui Kyong Chun describes as “the logic of what lies beneath,” the code here does not need added commentary (20). Indeed, in Andrew Galloway’s formulation, “Code is the only language that is executable… Code is the first language that actually does what it says” (165-66). The point of the poem is not something locked away in a safe: Combinatorial, the action of spinning the tumbler is the “tumbler” itself. Inscape digitalized. As in Gerard Manley Hopkins’s sonnet: kingfishers catch fire, dragonflies draw flame and, now, grasshoppers spin to the dictates of a digital prosody.

In JavaScript, these operations of reversal and transposition would be encoded as alerts and function calls. Of course, implementing the final function call in the second or third string of code would end the poem prematurely, contrary to the author’s playful intentions. Digitally, nonetheless, they represent the mental operations the observer must perform to make sense of what is being perceived: the poem’s gestalt. “Codework,” Raley informs us, “is a kind of object in that it puts on displays for the user-viewer, whose reactions and responses it then incorporates within its field of performance.” Noting that the brain is an information processing system and arguing that thinking itself is a form of computing, computationalists would be greatly interested in the poem’s construction of a grasshopper, inscape and all, through a cognitive process expressible in digital terms. Tracing “the core ideas of present day computationalism” to the seventeenth century, Matthias Scheutz establishes the deep historical roots of the notion of a mental calculator:

the notion of computation was intrinsically connected to the operations performed by mechanical calculators, on the one hand, and to cognitive processes using representations (such as calculating and reasoning), on the other. It was this link that eventually gave rise to the hypothesis that mind might be mechanizable. (5)

Somewhat counterintuitively, compiling the poem into a format approaching object code allows us to trace and “assemble” its patterning and execution more clearly, even though the assembly language of the computer is generally more comprehensible to the machine than to the human reader. Raley’s contention that “codework makes exterior the interior workings of the computer” finds a corollary in what “r-p-” accomplishes on the page. Function brings code to a level of near visibility that compares favorably to code poetry’s own.

The relatively orderly syntactic construction occurring between r-p-o-p-h-e-s-s-a-g-r and PPEGORHRASS continues in a different ordering format as we move from PPEGORHRASS to gRrEaPsPhOs. The progress towards an analog or plain text reconstruction is now expressed by odd and even integers, respectively increasing and decreasing almost entirely by increments of two:

Fig 5. Conversions for gRrEaPsPhOs.

As with PPEGORHASS, a bit of cross-switching at the pivot point makes the pattern even more orderly:

1 3 5 7     9 11     10 8 6 4 2

In his discussion of successful code works as “generative media,” Geoff Cox notes that in such productions “data is actually changed as the code runs” (Cox, McLean, Ward 7). Cox cites one poem in particular, in which “the ‘++’ and ‘- -‘ symbols are used to increment and decrement numbers” (7). A successful code work in its own right, Cummings’s poem, which consists of a playful rearrangement of letters on the surface, reflects an underlying generative matrix directing its operations.

This interweaving, along with the shuttling back and forth of numbers, recalls the movements of the Jacquard loom, whose mode of weaving threads into floral and leaf designs Lovelace felt bore a strong resemblance to the Analytical Engine’s own “weaving” of algebraical functions. Posing this analogy, Lovelace explained that the Engine required both “a scientific and emotive perception” for its “brilliance” to be fully understood: “We may say most aptly, that the Analytical Engine weaves algebraical patterns just as the Jacquard-loom weaves flowers and leaves” (Lovelace 696). In a bold proclamation concerning her analogy between the operations of the Analytical Engine and those of a loom, Essinger asserts that, “A strong case could be made that this sentence is the most visionary sentence written during the entire nineteenth century” (Ada’s Algorithm 169). Providing “a conceptual gateway” (Ada’s Algorithm 141), her poetic analogy involving a weaving process offers a practical application for visualizing what is going on in the poem’s computations. Thus, we find that the numerical expressions behind each encoding of the letters of “grasshopper” resemble the programming indicators of a loom-like operation. The hyphenation of “r-p-o-p-h-e-s-s-a-g-r” suggests the weaver’s plain stitch pattern:

Fig 6. Plain stitch pattern

The weaving analogy extends to PPEGORHRASS, which commands a shuttling motion, with its 8, 9, 10 and 2, 3, 4, 5 encodings mediated by the “cross-stitching” commands of its middle column (1, 11, 7, 6) and their final reversal. The under-and-over weaving pattern of the loom can be compared to the interlaced coding pattern of gRrEaPsPhOs, with the direction reversing at the cross-stitch point (11 9).

1 3 5 7     11 9     10 8 6 4 2

1 3 5 7     9 11     2 4 6 8 10

1 3 5 7 9 11

2 4 6 8 10

The integrating of odd- and even-numbered patterns in the last two rows is comparable to the lock-stitch function of weaving. Two threads, an upper and lower, are entwined or locked together in passing through the holes in a fabric.

Fig 7. Lock-Stitch Pattern of gRrEaPsPhOs. Elkagye. CC BY-SA 4.0. 11 Aug. 2011. Accessed 7 Oct. 2019.

A dynamic similar to the bobbin’s threading action is at work in the poem’s own weaving of letters. The final formation of “grasshopper,” digitally expressed by these odd and even integers, completes this entwining, locking process, once again illustrating the aptness of Lovelace’s comparison. The interlacing of “thinner” lower-case letters with “thicker” upper-case letters supplies a richer texture here as well. The semicolon at the end of “,grasshopper;” not only closes off the poem as a statement (in computer parlance) but also binds or “casts off” (in knitting parlance) the pattern of the poem. Cummings is enamored of this threading operation, a form of hyperbaton, or making a phrase or sentence discontinuous by inserting words into it. We can see this in the nineteenth poem of 95 Poems (1958):

Un(bee)mo

vi

n(in)g

are(th

e)you(o

nly)

asl(rose)eep

Here, two separate threads, comprising the sentences “Unmoving are you asleep” and “bee in the only rose,” are interwoven in fashioning the poem. Most readers have to unravel the two threads first to make sense of the poem’s dual statements and then re-encounter them in their original format. This process runs counter to what William Butler Yeats observed of the poet/reader relationship when he notes, “A line will take us hours maybe;/But if it does not seem a moment’s thought/Our stitching and unstitching has been naught” (“Adam’s Curse,” ll. 4-6). Early in his first draft of the poem, Cummings tried out an even more complex form of hyperbaton: “The up(now)gath(grass)eringhim(hop)self into(per)a.” Recognizing how his final version follows a coding pattern represented in this instance by PPEGORHASS and its digital format makes the stitching together a bit more manageable.

The three scrambled forms of “grasshopper” function on the order of a symmetric-key encryption in which a secret key is applied to information to change its content. The key can be composed of a number, word, or string of letters. The method of encoding can be as simple as shifting each letter by a number of places in the alphabet, the strategy employed to various degrees in “r-p-.” The permutations of “grasshopper” reflect a block cipher method of encryption first set forth by Claude Shannon. Also known as an iterated product cipher, “it carr[ies] out encryption in multiple rounds, each of which uses a different subkey derived from the original key” (Mathur 12). In their numerical equivalents, they can be

classified as iterated block ciphers which means that they transform fixed-size blocks of plaintext into identical size blocks of ciphertext, via the repeated application of an invertible transformation known as the round function, with each iteration referred to as a round. (Mathur 13)

At first glance, “r-p-” might strike readers as a random arrangement of letters, the sort that Shannon investigated in The Mathematical Theory of Communication. Through the probable sequencing order of natural language, Shannon demonstrated how a random selection of symbols could be processed from a zero-order approximation of a message through six stages to a final second-order word approximation (Shannon 13-14). Here, he helps his readers visualize a series of processes involving communication over a discrete noiseless channel, as found in teletypes and telegraphy. He employs these categories to illustrate the process of transformation.

Occupying the place of the title, “r-p-o-p-h-e-s-s-a-g-r” is more random than the other two subkeys, as it seems to signal a stochastic procedure in which the distribution patterns of the three forms and their surrounding text can be analyzed but their final expression may not be resolved at the end of analysis. The term constitutes a floating value as well, Max Nänny observing that the poem’s restless, agitated eleven letters

behave like grasshoppers in a bait box, wildly hop[ping] around in the poem, leaping lines, landing in the middle of a word (l. 5) or a sentence (l. 12). Even the title of the poem, I suggest, has hopped from its proper place to line 7 (‘The’) and line 14 (“,grasshopper;’), thus disguising the fact that the poem contains the fourteen lines of a sonnet. (134)

The handwritten draft of the poem partly bears out this assertion. In the draft, the one variation (“rporhessagr”) and final expression of “r-p-o-p-h-e-s-s-a-g-r” follow the segment “upnowgath / PPEGORHRASS / eringt(o.” With such a placement, “r-p-o-p-h-e-s-s-a-g-r,” Cummings’s final selection, was initially linked to (and, therefore, the missing subject of) the two articles (“aThe):”) that preceded it in the draft. What the draft reveals, however, is that the final expression hopped backwards from its original place in line 7 to occupy the place of the yet-to-be-determined title, left blank in the first draft. Linked to the indefinite article “a,” the final expression aptly reflects the very generalized notion of a grasshopper offered at the beginning of the poem. The definite article “The” directs our attention, in deictic fashion, to the ending (“.grasshopper;”). Capitalized, it would more properly stand as the title than does its lower-case expression. Like the last piece completing a jigsaw puzzle, the “The” directs our attention to the ending term “,grasshopper;” as the whole composition is drawn into focus.

Cummings may have moved this subkey from the middle to the top to serve not only as a title but as a more properly enigmatic expression of the poem’s subject matter and as a strategy of obfuscation. In contrast with the final form (expression) of the poem, “r-p-o-p-h-e-s-s-a-g-r” is the logical starting point for the digital-to-analog conversion performed here. Governing the most unreadable, unpronounceable elements of the poem, “r-p-o-p-h-e-s-s-a-g-r,” more randomly digitized, is closer in nature to a machine-language expression. It is more properly placed and can be considered as the determining, but missing, value for the middle section of the poem (beginning with line 7), which seemingly has no apparent governing subkey.

Appearing in the second and subsequent drafts and serving as the title in the poem’s published form, “r-p-o-p-h-e-s-s-a-g-r” is re-scrambled into PPEGHORASS and gRrEaPsPhOs; however, traces of it remain in the poem’s midsection, allowing it to serve as a subkey for “governing” the wildly chaotic movement of the section it underwrites. There is a partial alignment and expression of the first four letters in the bottom portion of the poem (G-r-a-S) and a fragment of some of the key letters in the top half “(o-p-e”):

Fig 8. Fragmentary reconstruction of “r–p–o–p–h–e–s–s–a–g-r

In relation to Shannon’s six categories of approximations to English, the term “r-p-o-p-h-e-s-s-a-g-r” falls somewhere between the first category, zero-order approximation (“symbols independent and equiprobable”), and the second category, first-order approximation (“symbols independent but with frequencies of English text”: e.g., “pop” and “sag”) (13).³ Citing observations from Moe and Etienne Terblanche (“Plotting” 134), Webster associates “pop” with the grasshopper’s explosive leap and “sag” with the effect of the insect’s landing on the grass (“Plotting” 134). These formations (p-o-p [8-7-9] and s-a-g [5-3-1]) are more descriptive of the actions of the grasshopper itself rather than serving as steps on the way to its orthographic reconstitution.

Unlike other poems in Cummings’s oeuvre, where randomization is never resolved, PPEGORHRASS marks the transition from Shannon’s fifth category to his sixth: from first- to second-order word approximations, or the appearance of word units. The phrase “who/a)s w(e loo)k” more properly fits category six, as the parentheses breaking up the word units distinguish it from full-fledged natural language. The subkey points to a more orderly pattern forming, with the block RASS approximating the word unit [G]RASS, a requirement of the first-order approximation. The blocks “upnowgath / PPEGORHRASS / eringint(o-” revert to category five, with recognizable but elided or separated word units.

Like the block units in PPE GORH RASS, the units in “[up][now][gath…ering][int(o]” must be shuffled around to comprise a more syntactically correct sentence: [now][gath…ering][up][int(o] or: [gath…ering][up][now][int(o]. With the same number of letters as the subkey “PPEGORHRASS,” the phrase “gathering up” mirrors the manner in which blocks comprising that subkey must be moved around reconstructed in order to reconstitute the name.

Finally, gRrEaPsPhOs falls roughly into the sixth category due to the “word transition probability” suggested by the lower-case formation, “g-x-r-x-a-x-s-x-h-x-s,” where “grass” is a highly probable construction. The digital expressions behind PPEGORHRASS and gRrEaPsPhOs reveal their apparent randomness as just that: apparent. They are functions of a programming script, the “software” through which the digital grasshopper is run.

Each round or iteration of the block cipher not only encodes the term “grasshopper” but also provides a governing subkey for the surrounding textual arrangement: “r-p-o-p-h-e-s-s-a-g-r” reflects the chaotic, asyntactical arrangement of its surrounding text. The syntax of the text surrounding PPEGORHASS reflects the relatively ordinal arrangement of the two blocks (PPE and RASS: 8, 9, 10 and 2, 3, 4, 5) comprising respectively the end and the beginning of the word. The mid part of the construction (GORH), with its out-of-sequence ordering (1, 7, 11, 9), defines the break in “gath…ering.” The ambiguous phrasing in the text surrounding PPEGORHRASS is also reflected in the digital expression (8, 9, 10 and 2, 3, 4, 5) outside the middle column. The adverb “now” can either modify the act of the observer (“who // ,a)s w(e loo)k/upnowgath”) or the action of the grasshopper (“upnowgath…eringint(o-“). An open question: Is one or the other placement “wrongwayroundfully” turned? A more regular syntactical arrangement would have “up” on the other side of PPEGORHRASS: “gathering up into.” The arrangements of the three digitally expressed columns index and govern the “word breakage” and asyntactical arrangement of the surrounding text. In doing so, they are the equivalent of function calls.

The third term or subkey, gRrEaPsPhOs, performs a similar function on the surrounding text. Its interweaving of the lower-case and upper-case letters of “grasshopper” programs the similarly constructed “rea(be)rran(com)gi(e)ngly” closely following it and reflecting (as does “rIvInG”) the subkey’s arrangement:

Indeed, the series of operations articulated earlier can be employed between the penultimate and last lines to describe the sorting process by which “gRrEaPsPhOs” becomes stated as “grasshopper;.” By inverting 11 and 9, we arrive at

1 3 5 7     9 11     10 8 6 4 2

thereby transforming the subkey to

.gRrEaPsPsOh;

Separating out the letters, we arrive at

g r a s s R E P P O h

Reversing the order of letters in the second column, changing the capital letters to lower-case, and supplying thereby some missing steps, we arrive at

.grasshopper;

Ensconced between a period and a semicolon, “.grasshopper;” is programmed and framed as a statement through grammatical markers to rest in place, literally and digitally, at this point.

We can apply the same program of separating (unraveling), sorting, and reversing functions—through which “gRrEaPsPhOh” was decrypted and transformed into “grasshopper”—to the phrase “rea(be)rran(com)gi(e)ngly” as well:

rea rran gi ngly (be) (com) (e)

(be) (com) (e) rea rran gi ngly

become rearrangingly

,grasshopper;

Like the codepoems it suggests, “r-p-” is, in Risam’s formulation, “variably accessible and inaccessible to readers, a function of their readers’ knowledge of programming languages and facility with poetry.” Of course, the range of “programming languages” available to readers trying to negotiate this poem was only in the process of being formulated and established when “r-p-” was first published. Readers had to parse and make sense of Cummings’s novel idiolect on their own. Like the language of the codepoem, this idiolect offers its own “visual aesthetics of … code on the page” as well as multiple “possibilities for interpretation […which] are fragmentary, requiring negotiation on these many fronts to appreciate and understand” (Risam). As an early manifestation and inadvertent anticipation of codepoetry, “r-p-” establishes the unconventional reading/decoding process of the format to come. 3. Perhaps, but…

One of this article’s two insightful reviewers took issue with the admittedly bold claim that “r-p-o-p-h-e-s-s-a-g-r” is a software program for running an executable file designated as grasshopper.exe. S/he remarked: “Not fully convinced that it is a software program. Can it be run on any other text?” As it turns out, yes, it can. In the same collection, No Thanks (1935), Cummings pro-grams Poem #48 (“Float”) to run in a similar fashion to “r-p-.” An examination of its first three lines illustrates the table of permutations guiding its operations.

His tribute to “the aristocrat of tap,” Paul Draper, begins with a verbal reenactment of his floating style:

floatfloafloflf

lloloa

tatoatf loat fl oat

Analyzing its arrangement, we compile the first three lines into a linear, digital format:

floatfloafloflf lloloa tatoatf loat fl oat

123451234123121 223234 5453451 2345 12 345

Parsing it by number blocks reveals some order among its otherwise seemingly arbitrary arrangements:

float     floa     flo     fl     f     lloloa     t     at     oat     loat     fl     oat

12345     1234     123     12     1     223234     5     45     345     2345     12     345

Avoiding for the moment the trickier issue of the sequence lloloa (223234), we can see that the poem’s “floating point” blocks of letters can be rejoined to form the original term, “float;” thus, the block numbered 1234 (floa) can be completed by joining it to 5 (t), twelve spaces across; the block numbered 123 (flo) by 4 and 5 (a t) thirteen and fourteen spaces across, and so on, as schematized below:

Fig 9. “Floating” Arrangements

With the formation lloloa (223234), we seem to have an arbitrary, “left-over” block of letters and numbers but, as the grid below demonstrates, the single f in the first chart does double duty in connecting l, lo, and loa to the blocks completing them (t at oat). (No doubt coincidentally, f recalls the f symbol representing a function.) This second set of blocks, used to complete the formations in the first grid, are re-employed in reverse order to complete the f lloloa formation:

[float     floa     flo     fl]     f     lloloa     t     at     oat     [loat     fl     oat]

[12345     1234     123     12    ]     1     223234     5     45     345     [2345     12     345]

Fig 10. f [unction] + lloloa + t at oat

A reversal similar to that noted in the discussion of PPEGHOHRASS occurs here as well. The t-at-oat-loat sequence completing the blocks in the first grid is now expressed in reverse order as loat-oat-at-t in the second grid. Like the grasshopper, Draper has pirouetted in the air: physically, alphabetically, and digitally. These numbers choreograph the tap dance steps described here. Bob Grunman describes a partial effect accomplished in this numerical encoding: “dropping the ‘f’ from ‘float’ in a series of repeating instances of ‘float,’ suggest[s] that the floating dancer floats away from himself” (84). The numeric play also “mathematizes” the turns and reverses of the performance. Compiling these first three lines into a digital format reveals not only the program in which they were stored but also the order of operations governing how the five letters were to be rearranged, digitally choreographed, and executed. Considered by itself alone, the digital format presented here would not make much sense to a human “reader,” being just as opaque as the binary coding of machine language directing the operations of a computer. Aligning it with the letters of “float,” an analogue expression, however, moves the format closer to the concept of pseudocode. It brings into clearer focus the algorithmic activity or instructions for the operations creating the syncopated “tap dancing” performed by the letter formations in these lines.

In further defense of my position that Cummings has fashioned a programmable code here, I called in an expert witness, programmer Steve Vogelaar, who was tasked with examining the above schema for “float” and determining its relevancy to programming code. For the more technically proficient, refer below to his encoding of “float.”⁴ In plain prose, he finds:

The poet’s disassembly and reassembly of the keyword “float” lends itself to programming code very efficiently. The code that would parse the first two lines works very well with standard “LOOP” programming using an incremental or decremental counting system. These “LOOPS” concatenate the keyword to itself, either adding an incrementing number of characters or adding a decremented number of characters each time until the loop ends, depending on the type of “LOOP” and the character parameters of each loop.

The third line’s first word is also coded easily with another “LOOP.” The last three words (“loat”, “fl” and “oat”) being singular and not concatenated pieces of the keyword, can easily be broken apart and output in pieces using another coding tool that allows us to use “character counts” and their numerical position within the word to pick and choose the parts we need.

By using “LOOPS,” character counting, and placement principles, it is possible to recreate the permutations of the keyword output as the poet has done. The idea behind the “pseudocode” was to show that the poet’s disassembly of the keyword could be digitally programmed so that any word could be input to keyword, and the output of the poem would adhere to the same order of permutations no matter how many letters in the word.

It would appear that the poet, although not a computer programmer, has used future computer programming logics, whether he realized it or not. Or did the tech industry borrow from our poet? A discussion for another day.

To his closing comments, we can pose the question of agency. If Cummings is not consciously coding this poem in whatever nascent fashion, and it is highly improbable that he would be doing so in 1935, then where is the locus of agency here? Is a coding process yet to be formally articulated already in progress and operant here? As in the leaping grasshopper and thus in the graceful tap dancer as well, does inscape lend itself to digitalization? Is the outward expression of an inward, dynamic self-programmable?

The numerical sequence at work here is thus comparable and translatable to the software running a program. For Wendy Hui Kyong Chun, software constitutes “the logic of what lies beneath” (20). Here, the operations of that logic play out in a more visible fashion than is customary in Chun’s depiction of software. It is more in line with Galloway’s concept of code as “a machine for converting meaning into action” (166). Performative rather than simply descriptive, Cummings’ poem enacts its subject. It is encoded in what Chun describes elsewhere as “a very special language” (22). She offers Galloway’s pronouncement, cited earlier, in driving home her point: “Code is the only language that is executable….code is the first language that actually does what it says” (165).

Towards the end of the poem, the equation “spun=flash,” is an allusion to the flash routine Draper perfected, an acrobatic fusion of ballet and dance, a style dance critic John Martin praised in noting that Draper had “chosen tap dancing as a medium for art rather than for hoofing.” The critic goes on to note: “The arms served a definite function in the guidance and propulsion of the body, there was a kind of vertical unity between the feet and the rest of the body” (Cited by Hill, 191). This encoding of tapping and action can be expressed both analogically

;d

;;a:

nC.eda:Nci;ddaanncciinn

(GIY)

and digitally

;1 ;;2: 34,512:34 6; 112233446677 (8 9 10)

Not surprisingly, given our programmer’s analysis, a strict set of numerical, “combinatorial” operations lies behind the poem’s fanciful, seemingly arbitrary arrangements. Grammatical symbols (semi-colons, colons and commas) are employed as beat markers (comparable to the use of the ampersand in denoting beats). The poem’s repetition and spacing of letters serve the same functions as movement indicators found in tap dancing protocols, as set forth by Sara Pechina (101):

Digitizing the poem demonstrates how Cummings has followed a very methodical process in creating the poem’s special effects, here the choreographing of dance steps.

As it turns out, a strict set of mathematical, “combinatorial” operations lies behind the poem’s fanciful, seemingly arbitrary arrangements, particularly helpful in sorting through the more jumbled arrangement of its middle section. A similar logic of transposition, shift, and reversal governs the order of operations in both works. In contemporary terms, we would identify the relationship between these two poems as involving a function call. Here, a called (or inner) function defining one poem/program’s operations is invoked by another’s calling (or outer) function. Given that “r-p-o-p-h-e-s-s-a-g-r” is numbered thirteenth and “float” forty-eighth in the collection, we might label “float” as the subroutine of a program initiated by “r-p-o-p-h-e-s-s-a-g-r.”

4. Is It in the Cards?

We can also find it instructive to move from digital expressions underlying the operations of the poem to an overview and appreciation of its “visual aesthetics.” In describing how “punctuation marks … highlight the structural disorganization of the poem,” Eva Maria Gómez-Jiménez advocates the same strategy: “the text has to be looked at from a distance and as a single object, rather than as a poem to be read aloud” (201). As an encryption program, the poem is laid out visually as though it had been written with a Cardan grille, a device used for encrypting messages and dating back to the Renaissance. The Cardan grille had variously sized, randomly placed rectangular apertures cut into it. The encrypter placed the card over a sheet of paper, writing the secret message in the apertures. Removing the card, the encrypter would fill in the blank spaces with an innocuous-sounding message:

TRIUMPHS APPLY TO THOSE WHO ARE CAPABLE 
OF ADJUSTING TO THE CHALLENGES SOME FOLKS 
CREATE TO MEASURE SUCCESSFUL AGE 
INHIBITED BEHAVIORS EVERYONE ENCOUNTERS 
DUE TO DEFICITS IN THIS COGNITIVE RECALL OF 
SELECT INFORMATION OF CARDINAL ORDER 
MIMICKING JUST A RANGE OF A NONCE USAGE

Once this message was sent, the recipient at the other end would have the same card and would place it over the message, the apertures now revealing the disguised message:

Fig 11. The secret message displayed by employing the Cardan grille.

The analogy to Cummings’s poem seemingly stops short here, as its blank spaces are not filled in. Counting the mostly scanty lines of the poem, however—fourteen of them—we recognize that there is an underlying, invisible “content” or structure here: the sonnet, though it is disguised by the poem’s asymmetrical, spotty arrangement. Imagining a Cardan grille placed over the poem reveals a coded “message” revealing its sonnet structure. Here, the coding process has been turned inside-out, the text message on the surface and the hidden message concerning the poem’s form encoded. Had the blank spaces been filled in with “filler” content, the poem would have been more easily recognized as a sonnet; however, we might simply be left with an innocuous statement about a grasshopper, with none of the dynamics involved in deciphering the poem. We would certainly be missing the scattering effect achieved by encrypting a message like this…

The Cardan grille application also reveals that the poem is not a simple cipher; even if we imagine we are looking through the apertures of a card placed over the “sonnet content” of the poem, we would have to deal with digital (coded) as well as analog text. The digital text is comprised of the three subkeys of “grasshopper,” surrounded by analog (though scattered) text. Plain text or source code (the readable characters of data in a file) and object code (machine-readable text) are in tension here. The unconventional use of “dislocated” commas, periods, semicolons, colons, and an exclamation point scattered throughout the text also suggests a vacillation between the analog and digital. “How have they been re-coded?” Readers must wonder. The three forms can likewise be viewed as supplying an algorithmic formula for calculating functions. Indeed, each one provides instructions to its surrounding text as the poem moves—is executed—from an initial state through a series of successive states to terminate in an ending state (“;grasshopper;”). The three subkeys govern the sequence of operations that define the software program being run.

Another card must be dealt in our investigation. To understand Cummings’s codework more completely, we must move from the Cardan grille analogy to one represented by the Jacquardian punch card, a movement in effect from the analog to the digital. We have established how the digitized key word (“grasshopper”) reveals the poem’s programmatic, digital operations as reflected in the variant forms/functions of its three expressions. Unlike the simple coding provided by the Cardan grille, the encoding of punch cards lends itself more readily to digital operations, the critical insight Lovelace made in connecting the operations of the Analytical Engine to those of Jacquard’s loom: “These cards contain within themselves the law of development of the particular function that may be under consideration, and they compel the mechanism to act accordingly in a certain corresponding order” (Web 141).

________

For the purpose of illustration, the poem can be likened to the punch cards driving the operations of Jacquard’s loom and proposed for the operations of the Analytical Engine. The poem’s three sections, coded through r-p-o-p-h-e-s-s-a-g-r / PPEGORHRASS / gRrEaPsPhOs, can be compared to the function calls of a series of three punch cards. As Babbage informs us of his own Engine, it “first computes and punches on cards its own tabular numbers” (Jacquard’s Web 91). Converting these alphabetical formulations to digital expressions—Babbage’s “tabular numbers”—has allowed us to perceive more clearly those operations at work. Lovelace’s explanatory description of the Engine as weaving algebraical patterns in the same fashion as the Jacquard loom wove designs offers for our purposes a further means for visualizing the operations of Cummings’s poem. What might the poem look like executed on a Jacquard loom?

To begin with, the poem would need a warp through and upon which its pattern could be woven. The default sonnet here—implicit rather than explicit—functions as the warp or frame of the poem, comparable to the longitudinal, stationary threads through which the transverse weft material is drawn in the loom’s over-and-under process of creating a pattern. The weft or “filler thread” in this analogy comprises the words and phrases woven into the sonnet frame’s “warp.” The digital cross-stitching noted above also appears at the phrase or word level at several points in the poem. As noted earlier, the hyphens in “r-p-o-p-h-e-s-s-a-g-r” function like stitching to connect the letters. The interlacing of lower-case and upper-case letters in the formations “rIvInG” and “gRrEaPsPhOs)” suggests the up-and-down, over-and-under motions of a loom’s shuttling process. Finally, the over-and-under action of the loom is suggested in the interlacing of “rearrangingly” and “become”: “rea(be)rran(com)gi(e)ngly.”

In their draft books, weavers would sketch out patterns to be used in fashioning the punch cards. They would visualize the threaded hooks’ operations and, as in the far right side of the figure below, would draw tables indicating the placement of the cards’ holes (the zeros in the various grids). If the rod passed through a hole in the card, the equivalent of “on,” its hook was engaged and contributed to the pattern’s construction. Rods deflected by the card, the equivalent of “off,” could not pass through and thus were not part of that phase of the weaving process. A binary code!

Figure 12. A page from a weaver’s draft book. Unknown. Weaver’s Draft Book (1805). Creative Commons. Accessed 7 Oct. 2019.

These notations, in turn, would be transferred over to punch cards that would direct the loom’s operations:

Figure 13. Representation of a Cardan grille

Imagining that the poem was “woven” from an underlying diagram that determined the weft and warp as it appears on the page, we can think of the blank spaces of the poem, the missing elements of a default sonnet structure, as representations of those places on the punch card where the threaded hook was prevented from penetrating the “page.” The letters of the poem have been “threaded” into the page as directed by the holes in the card. In Jacquard’s system, each card determining one row of design. A “punched in” version of the poem mimics the formatting found in these cards:

Fig 14. Illustration of the poem as the product of a punch card.

Each subkey, with its encoding of a digital arrangement, is analogous to the operations of a punch card governing the poem’s design. Thus, just as one card in Jacquard’s scheme deterrmined one row of design, so does each subkey determine the rows of the poem’s design. Like the code poets’ own productions, Cummings’s poem mixes analog and digital (coding) elements and displays them on the surface of the text. Where “r-p-” differs from Jacquard’s setup is that the poem includes both the pattern and also the cards producing that pattern (the three subkeys) in its design. In effect, it thereby enacts what Raley qualified earlier as “the art of the code, in which the code used to produce the work seems to infiltrate the surface, the former domain of natural languages.”

Interestingly enough, there has been some renewed interest lately in the connections between weaving and coding. In an article entitled “Weaving Paved the Way for Computer Coders,” Emiliano Rodriguez Mega and Lexi Krupp describe a workshop designed by Francesca Rodríguez Sawaya and Renata Gaui: Weaving to Code, Coding to Weave. As visual designer Sawaya explains, the goal of the workshop is to bridge the divide between crafters and coders. Observing that “weaving paved the way for coders,” she notes parallels between the two operations: “Computers talk about binary systems…We talk about ones and zeros but if you go back to weaving, you go up or down with your threads, and that’s basically going one… zero… one… zero… one… one… zero… one.”

Operationally, via the interactions between microfragments, words, and phrases, “r-p-” and its loom-like operations seem patterned after those of the Analytical Engine, as described by Lovelace:

It must be evident how multifarious and how mutually complicated are the considerations which the working of such an engine involve. There are frequently several distinct sets of effects going on simultaneously; all in a manner independent of each other, and yet to a greater or less degree exercising a mutual influence. (Lovelace 710)

The “sets of effects” in this instance are the operations of the three subkeys, which, operating independently, nonetheless exert a mutual influence on each other as the poem proceeds. The over-and-under/under-and-over permutations of the poem’s weft require, to borrow Nänny’s formulation, a “reading process [which] follows lines of motion that provide a diagrammatic icon of the elusive, haphazard jumps and flights of a grasshopper” (134). These lines of motion result in a complicated pattern of reading, at times shuttling right to left and then reverse shuttling left to right. While computationalists’ claims that the mind might be mechanizable still hang in the balance, “r-p-” demonstrates how words, properly placed and displaced, can make the mind move in a mechanized, loom-like fashion.

Demonstrating his own designs on the poem, Nänny finds a “poempicture” outlined therein. A veritable connect-the-dots representation of its outline, Nänny’s static representation is reductive and falls far short of what Cummings has actually achieved (view Nänny’s design: https://www.e-periodica.ch/digbib/view?pid=spe-001:1985:2#138). The mind at work here, an Analytical Engine in its own right, deserves a better rendering of its operations. In her translation of “Sketch of the Analytical Engine,” Lovelace describes how the Engine itself does not think but rather executes “the conceptions of intelligence” (689). Referring to the Jacquard loom, she likens these conceptions to the information punched and programmed into cards that direct the operations of the loom. Receiving “the impress of these conceptions,” the cards would then “transmit to the various trains of mechanism composing the engine the orders necessary for their action” (689). What “r-p-” demonstrates is a mind in the process of programming and transmitting a code for constructing the dynamic operations of its subject. Code poetry in motion, “r-p-” is a software program for running an executable file designated as grasshopper.exe.

Postscript

There is much speculation these days about what we might achieve by developing a mind/machine interface. In a recent essay, “Thought-Reading Machines and the Death of Love,” Jason Pontin references Ludwig Wittgenstein’s thought experiment, which begins with the proposition that everyone has a box with something in it called a “beetle.” Pontin continues:

Denying the possibility of private language, the philosopher wrote, “No one can look into anyone else’s box, and everyone says he knows what a beetle is only by looking at his beetle.” Wittgenstein meant that we learn a word by observing the rules governing its use, but no one sees another person’s beetle: “It would be quite possible for everyone to have something different in his box,” or nothing at all. An apparently intractable fact of life is that our thoughts are inaccessible to one another. Our skulls are like space helmets; we are trapped in our heads, unable to convey the quiddity of our sensations.

Pontin proceeds to discuss the work of researcher Mary Lou Jepsen, whose company, Openwater, is undertaking a project to holographically record and analyze the image patterning and very thought processing of individuals, a project straight out of a Black Mirror episode. In effect, Pontin maintains, she “wants to show me the beetle inside your box, and you the beetle inside mine.”

While it will be interesting to see what emerges from her investigations, the imagery will most likely fall short of what Cummings offers in “r-p-o-p-h-e-s-s-a-g-r,” a poem that operates like a three-part series of punch cards directing—Analytical Engine-style—the operations of the Jacquard loom. The poem should be imagined spun from the shuttling action of hooked rods, spinning cog wheels, and the clatter of leaping typewriter keys: a digital grasshopper woven out of software and threadware into the algebraic patterns of Ada Lovelace’s flowers and leaves.

Footnotes

1. Further references to “r-p-o-p-h-e-s-s-a-g-r” as the poem’s title will be abbreviated to “r-p-“

2. Poets perform a similar charting in writing the sestina, with its juggling of numerically marked end-words from stanza to stanza (e.g. 1, 2, 3, 4, 5, 6; 6, 1, 5, 2, 4, 3, and so on.) Quite aptly, practitioners of this form were sometimes called jongleurs (jugglers).

3. The following is excerpted from Claude Shannon’s The Mathematical Theory of Communication. He writes: “To give a visual idea of how this series of processes approaches a language, typical sequences in the approximations to English have been constructed and are given below. In all cases we have assumed a 27-symbol ‘”alphabet,'” the 26 letters and a space” (14). The following are Shannon’s desriptions and examples of the six different levels that comprise the “Series of Approximations to English” (14-15).

1. 1. Zero-order approximation (symbols independent and equiprobable)
   XFOML RXKHRJFFJUJ ZLPWCFWKCYJ FFJEYVKCQSGHYD QPAAMKBZAACIBZLHJQD.
2. 2. First-order approximation (symbols independent but with frequencies of English text).
   OCRO JLI RGWR NMIELWIS EU LL NBNESEBYA TH EEI ALHENHTTPA OOBTTVA NAH BRL
3. 3. Second-order approximation (digram structure as in English).
   ON IE ANTSOUTINYS ARE INCTORE ST BE S DEAMY ACHIN D ILONASIVE TUCOOWE AT TEASONARE FUSO TIZIN ANDY TOBE SEACE CTISBE. (13)
4. 4. Third-order approximation (trigram structure as in English).
   IN NO IST LAT WHEY CRATICT FROURE BIRS GROCID PONDENOME OF DEMSTURES OF THE REPTAGIN IS REGOACTIONA OF CRE.
5. 5. First-order word approximation. Rather than continue with tetragram,…, n-gram structure it is easier and better to jump at this point to word units. Here words are chosen independently but with their appropriate frequencies.
   REPRESENTING AND SPEEDILY IS AN GOOD APT OR COME CAN DIFFERENT NATURAL HERE HE THE A IN CAME THE TO OF TO EXPERT GRAY COME TO FURNISHES THE LINE MESSAGE HAD BE THESE.
6. 6. Second-order word approximation. The word transition probabilities are correct but no further structure is included.
   THE HEAD AND IN FRONTAL ATTACK ON AN ENGLISH WRITER THAT THE CHARACTER OF THIS POINT IS THEREFORE ANOTHER METHOD FOR THE LETTERS THAT THE TIME OF WHO EVER TOLD THE PROBLEM FOR AN UNEXPECTED.

4. Encoding of “float”:

[poemword] = “float”

/* NOTE POEM LINE 1 GENERATION */

[var_decrement] = 0

[var line1#] = 1

[var_line1_text] = “”

LOOP WHILE [var_decrement] < [poemword]length

[newword] = [poemword]character#[var_line1#] to [poemword]character#([var_line1#] -[var_decrement])

[var_line1_text] = [var_line1_text] + [newword]

[var_decrement] = [var_decrement] + 1

END LOOP

/* NOTE – POEM LINE 2 GENERATION */

[var_increment] = 0

[var_line2#] = 2

[var_line2_text] = “”

LOOP WHILE [var_increment_line2] < ([poemword]length – [var_line1#])

[newword] = [poemword]character#[var_line2#] to [poemword]character#([var_line2#] + [var_increment])

[var_line2_text] = [var_line2_text] + [newword]

[var_incremen] = [var_increment} + 1

END LOOP

/* NOTE – LINE 3 WORD 1 GENERATION */

[var_increment] = 0

[var_line3_word1] = “”

LOOP WHILE [var_increment] < ([poemword]length – [var_line1#])

[newword] = [poemword]character#([poemword]length – [var_increment]) to [poemword]char-acter#[poemword]length

[var_line3_word1] = [var_line3_word1] + [newword]

[var_increment] = [var_increment] + 1

END LOOP

[var_line3_word1] = [var_line3_word1] + [poemword]character#[var_line1#]

/* NOTE – LINE 3 WORD 2 GENERATION */

[var_line3_word2] = [poemword]character#([var_line2#] to [poemword]character#[poemword]length

/* NOTE – LINE 3 WORD 3 GENERATION */

[var_line3_word3] = [poemword]character#[var_line1#] to [poemword]character#[var_line2#]

/* NOTE – LINE 3 WORD 4 GENERATION */

[var_line4_word4] = [poemword]character#([var_line1#] + [var_line2#]) to [poemword]character#[poemword]length

/* NOTE – POEM LINE 3 GENERATION */

[var_line3_text] = [var_line3_word1] + ” ” + [var_line3_word2] + ” ” + [var_line3_word3] + ” ” + [var_line3_word4]

/* DISPLAY POEM */

DISPLAY [var_line1_text]

DISPLAY [var_line2_text]

DISPLAY [var_line3_text]

Works Cited

• Chiaverini, Jennifer. Enchantress of Numbers: A Novel of Ada Lovelace. Dutton, 2017.
• Chun, Wendy Hui Kyong. Programmed Visions: Software and Memory. MIT P, 2011.
• Cox, Geoff, Alex McLean, and Adrian Ward. “The Aesthetics of Generative Code.” Generative Art 00, international conference, Politecnico di Milano, Italy, 200. www.generative.net/papers/aesthetics/. Accessed 10 Mar. 2019.
• Cramer, Florian. Digital Code and Literary Text. Research Gate, 2019. https://www.netzliteratur.net/cramer/digital_code_and_literary_text.html Accessed Sept. 17, 2019.
• Cummings, E.E. Complete Poems (1904-1962). Edited by George J. Firmage, Norton, 1991.
• Espen, Aarseth. Cybertext: Perspectives on Ergodic Literature. Johns Hopkins UP, 1997.
• Essinger, James. Ada’s Algorithm: How Lord Byron’s Daughter Ada Lovelace Launched the Digital Age. Melville House, 2014.
• —. Jacquard’s Web: How a Hand-Loom Led to the Invention of the Information Age. Oxford UP, 2004.
• Friedman, Norman. E. E. Cummings: The Art of his Poetry. Johns Hopkins UP, 1960.
• Galloway, Andrew. Protocol: How Control Exists after Decentralization. MIT P, 2004.
• Gómez-Jiménez, Eva Maria. “Unconventional patterns in the experimental poetry of E. E. Cummings: A stylistic approach to punctuation marks.” Language and Literature, vol. 26, no. 3, 2017, pp. 191-212. SAGE, doi:10.1177/0963947016686606.
• Grunman, Bob. “The Importance of Technical Innovation in the Poetic Maturation of Cummings.” Spring, no. 13, 2004, pp. 74-89.
• Harwood, Graham. “London.pl.” runme.org/feature/read/+londonpl/+34/. Accessed 6 Nov. 2018.
• Hayles, N. Katherine. “Print is Flat, Code is Deep: The Importance of Media-Specific Analysis.” Poetics Today, vol. 25, no. 1, 2004, pp. 67-90. Project Muse, doi:10.1215/03335372-25-1-67.
• Hill, Constance Valis. Tap Dancing in America. Oxford: Oxford UP, 2014.
• Leech, Geoffrey. Language in Literature: Style and Foregrounding. Routledge, 2013.
• Lovelace, Ada. “Notes by the Translator.” Sketch of the Analytical Engine invented by Charles Babbage. By L.F. Menabrea. Scientific Memoirs, vol. 3 (1842). London: Richard and John E. Taylor, 1843, pp. 691-731.
• Lovelace, Ada, translator. Sketch of the Analytical Engine invented by Charles Babbage. By L.F. Menabrea. Scientific Memoirs, vol. 3 (1842). London: Richard and John E. Taylor, 1843, pp. 666-690.
• Mathur, Aditya. Private Key Encryption and Network Security. Uploaded 27 Oct. 2012, scribd.com/document/111275631/Network-Security. Accessed 19 Nov. 2018.
• Mega, Emiliano Rodriguez and Lexi Krupp “Weaving Paved the Way for Coders.” New York University Journalism Projects. https://nyujournalismprojects.org/pixel/weaving-paved-the-way-for-computer-coders/. Accessed 7 Oct. 2019.
• Montfort, Nicholas. Exploratory Programming for the Arts and Humanities (London: The MIT Press, 2016), 92.
• Moe, Aaron M. “Autopoesis and Cummings’ Cat.” Rupkatha Journal on Interdisciplinary Studies in Humanities, vol. 3, no. 1, 2011, pp. 110-20.
• Nänny, Max. “Iconic Dimensions in Poetry.” On Poetry and Poetics, vol. 2 of SPELL: Swiss Papers in Language and Literature, edited by Richard Waswo, Gunter Narr, 1985, pp. 111-135. E-periodica, doi:10.5169/seals-99839.
• Pechina, Sara. “Steps in Time: An Exploration of Tap Dance Education.” 8/29/2016. https://digitalcommons.wku.edu/cgi/viewcontent.cgi?article=1657&context=stu_hon_theses
• Pontin, Jason. “Thought-Reading Machines and the Death of Love.” Wired Magazine, 16 Apr. 2018, www.wired.com/story/ideas-jason-pontin-openwater. Accessed 30 Jun. 2019.
• Raley, Rita. “Interferences: [Net-Writing] and the Practice of Codewriting.” Electronic Book Review, 8 Sept. 2002, www.electronicbookreview.com/thread/electropoetics/net.writing. Accessed 30 Jun. 2019.
• Risam, Roopika. “The Poetry of Executable Code.” Jacket 2, 5 Apr. 2015, www.jacket2.org/commentary/poetry-executable-code. Accessed 30 Jun. 2019.
• Tartakovsky, Roi, “E. E. Cummings’s Parentheses: Punctuation as Poetic Style.” Style, vol 43, no. 2, 2009, pp. 215-247. Proquest, doi:220126366.
• Webster, Michael. “Cummings’s Silent Numerical Prosody,” Media Inter Media: Essays in Honor of Klaus Cluver, vol. 3, 2009, pp. 423-38. Amsterdam: Brill and Rodopi. 426. doi: https://doi-org.ezproxy.library.uvic.ca/10.1163/9789042028432_021
• —. “Plotting the Evolution of a r-p-o-p-h-e-s-s-a-g-r.” Spring, vol. 20, 2013, pp. 116-42. https://www.academia.edu/29917786/Plotting_the_Evolution_of_a_r-p-o-p-h-e-s-s-a-g-r