On LaTeX, or My System of Sanity Maintenance as a Graphic Designer

An essay by Janis Vivian Media Lago

    Seldom do we stray anymore from the path of immediacy in graphic design: a plurality of changes run primarily through the mouse to be inflicted, inscribed, onto a digital surface limited in resolution only by available memory, layered for convenience, obstinate that everything be done as unprogrammatically as possible, the belief being that the designer, as a digital artist, is more "artist" than are "digital" and therefore should only have to endure the gritty internals of composition in place of interaction with the underlying machine framework, but this motive and its intentions merit interrogation.  The keyboard temporally predates the mouse by decades even within computing, and upon its introduction, a full what-you-see-is-what-you-get system was still yet to manifest given the difficulties overcome by software engineers maximizing potential within computational limits of the day, often strict and arbitrary even within workstations optimized for the precise purpose of graphics.  Therefore, early digital design often takes a limited form not unlike Knuth's famed TeX typesetting system, an ingenious conglomerate of all that was possible to execute in raster (or close) page design at the time, nothing immediate, everything entirely ascertained through carefully-timed printouts to ensure a lack of personal error not detected by the program itself with its abundance of backslashes, braces, and currency symbols.  TeX, of course, originally ran on a terminal, the technology most readily accessible to the academic wishing to make a preprint containing mathematical formulae but unwilling to settle for the displeasure of including hand-drawn ones within a given journal, and it included a related technology, METAFONT, used for the rendering of raster glyphs optimized and individualized for each possible printer resolution at which a document could ultimately be produced, but it was distinct from later desktop publishing software insofar as it only necessarily required a text editor into which it could be written, entirely free software to compile it, and a way to view and verify the formatting of the result.  It was not PageMaker, QuarkXPress, or Microsoft Word, instead stemming ultimately from the culture of Incompatible Timesharing System and WAITS, also the origin of Emacs, my preferred environment for editing TeX and its relatives.  It may be inevitable to wonder whether any of this may be of relevance to anyone in the design world when the clickable menu-based environment presents an immediate realm of possibility and prevents the designer from indulging in the framework by which the image is generated, but nonetheless, I seek to present a case for a technology primarily restricted to academic publishing as an artform unto itself.

    I will primarily refer to LaTeX herein because it presents an interesting contradiction: a typesetting language layered carefully onto its preceding TeX with the express intention of separating content from presentation, as is often the norm in web design, seems far from apt for solely artistic pursuits given that it intentionally prioritizes the content itself above its exact look or position on the page, yet it presents complete facilities for page design through a wide variety of extensions for graphics, text positioning, page sizing, et cetera, so it has become my language of choice from my very first introduction through deciphering org-mode export backends (indeed, this turned out to be much more of a rabbit hole than I initially anticipated, hence the existence of this piece).  I, however, was initially oblivious to this contradiction and, not understanding the ethos surrounding LaTeX, challenged myself immediately to put it to use for highly context-dependent visual art bound to virtual pages and exported direct-to-PDF (as I still do today), a restricted shorthand for a concatenative language at a far lower level of page design, yielding interesting results in due time, especially as I began to work with PGF/TikZ, which is, of course, not a drawing language, but a way to describe graphics briefly and eloquently.  Through such processes, I found I could compose and align without any trouble from auto-alignment clicked with a mouse, specify features of text without crawling relevant menu after relevant menu until hallucinations of irrelevance of it all and laziness began to overtake my vision, precisely coordinate coordinates with elaborate iterations or transformations, and generally discover alternative solutions whenever one aspect of the code worked in any way other than perfectly, a viable strategy for genuine graphic design resulting in the process.  The layers of abstraction separating me from low-level PostScript framed then and still frame design in LaTeX as a compositional exercise not dissimilar to ordinary writing, with text directly integrated and all components designed for immediate legibility, a statement unspeakable of a program as complex as, say, Adobe Illustrator.

    Illustrator, with which I had a brief flirtation about three years ago now, frustrates me to no end because, while expressly intended for vector design, it is just as ancient a technology as LaTeX or PostScript executed in the absence of interfacing with the intensely mathematical nature of vector graphics; all things Illustrator, rather, are expressly bound to measurements which must, for the most part, be measured manually with keyboard-and-mouse mechanics and alignment intended for us to click, inscribe, draft as if still bound to the pen, paper, ruler, compass, and so on.  This is not to say I wish to ascribe or pin dishonor onto pen-and-paper design approaches because there is one important detail in this approach which transfer over to digital graphic design and indeed does despite a lack of transparency in the forms of programs, the aspect of geometric construction.  To draw from the honorable field of vexillology, designers of flags wishing to codify their creations into law undergo scrutiny, executing a complex process of description applicable to domains of shape and color alike such that the flag is reproducible in manufacturing from merely a written account of its proportions and other visual features, essentially known in idea, but made relative to cataloging systems such as Pantone when necessary.  Consider, then: project files for a page or vector design are likewise a description from which a design may be reproduced, but to the erudite analog designer of pre-computational yesteryear such instructions - or declarations, depending on one's outlook - are inconceivably exact down to units as small as a single 1114112th of an inch, as found in the TeX system about which I write, but making the compromise between human and computer standards of legibility proves itself inevitable and complex alike whenever it appears - it always does - and, for we are already embroiled in a situation pseudo-distanced from its mathematical underpinnings, we find the format tends toward the machine's whims, just as it did in days of restricted memory and computational power, indecipherable and far from transparent.  The former directness of design in its physicality was now lost within encoding, architecture, and a litany of other uniquely digital factors, but was what-you-see-is-what-you-get a one-size-fits-all solution for accessing the data therein, or was everything to be considered write-only, conjugation tables which the program must simply memorize for their irregularities, often concealed in cryptography or compression far more complex than we can penetrate with pen-and-paper or M-x decipher RET.  Alas, our formats, our webs of interconnected formatting information, are caught up in a web of our own illiteracy, their behaviors undocumented to the general public.

    Now, consider also another domain, one indulged in out of necessity and-or under the pretense of serious graphic design far too often by those who have by and large never used any sort of desktop publishing or typesetting program otherwise.  Microsoft Word, saving near-inevitably in a proprietary format (not even considering the difficulties with the proprietarity, to which many arguments regarding or against Word have been made in a continuous stream since its inception) predominantly binary, with dependencies on behaviors of particular versions of Word, which create a dependency on Word itself for backwards compatibility, and the interface presented by Word upon startup by an unsuspecting user seeking to write a paper or design a flyer is, to say the least, obtuse to its own detriment, discombobulated and complicated by a network of, again, buried graphical menus - click, click, click, click - perpetually mouse-governed in a loop of formatting, a Proustian death sentence for a modern Sisyphus.  My experience with Word and its numerous imitators - all with their own sets of difficulties quite often rooted in the licensing I have promised not to discuss in-depth here due to my relative lack of expertise, limiting myself to the discursively coverable within a gamut of reason - was never even dotted, much less crossed, with pleasantries when formatting papers throughout secondary school with strict limitations of format, pervasive to the extent that the gestural sequence resulting in Times New Roman, twelve-point-font, double-spaced, and one-inch margins became conceptually automatic, but the gesturality required me to turn away from my keyboard, never divorced the information itself from my instructors' expected formatting looks or desired textual lengths demarcated with surveys in units dictated by formatting standards (themselves seemingly equal in their arcane complexities to the difficulties of redefining the metric kilogram without any dependency on the International Prototype whatsoever, although far less rigorous), always took too much and required the use of a system of troubled templates not necessarily fitting.  "No title page is required," some might have said.  "Page numbers do not need to match the APA," others uttered.  The documents themselves never ended up pleasing the eye particularly, their alignments uncleanly and their hackneyed structures irregular and warped to fit an assortment of needs often too broad for Word itself, all in pre-installed fonts seemingly popping into being with every wish I had to edit the format of my page numbering, my footnotes, my bibliography, and thus the idea that the document should consist of one font all the way through as a colloid of glyphs made so even it may as well be one surface received little in the way of respect.  Therefore, I would neither design nor typeset in Word: as a program, it seems a flimsy replacement for LaTeX when one discovers the latter.

    One commonality between Microsoft Word's ideal domain of the many-paged, sprawling paper and Adobe Illustrator's preferred productive space of the vector graphic is that both areas of work, despite the significance in the divergence of prioritizing either linguistic or abstracted-visual information, ultimately convey a message in some regard, read in by the viewer-examiner in manners distinct from the comprehension of explicit speech, immediacy not resulting from the nature of the interaction itself but rather purposefully composed features of the visual information, be it in a bolded word or the centering of an element, in the angularity of a tilted vector or the typographic offsetting of a text block away from its companions through a distinguishing indentation.  While this is common knowledge to its users and I will make no serious attempt to evangelize, LaTeX covers both domains while ensuring that formatting is adaptable, that no renumbering of a referenced example will be affected by the addition of another before it, offsetting the enumeration and vanishing the need to look for all examples thereof within a paper and alter their numeric values, which, from the perspective of a linguist whose example (re)numbering is presently automatic and who has to keep her examples numbered consistently without any duplicates, is utterly glorious, for labels and refs in tandem render the document presentable and adherent to a format without much effort on my own part, a fine-tuned weighing scale - Lady Justice's balance, if you will - proving the masses of intentionality and automatic convenience entirely even in their prioritization by both myself and the program, a plethora of convenient resources available, a plurality of editors presenting themselves to each potential user, a series of opportunities, even preferentially chosen engines with reproducible results individualized to each, all of which remain well-loved even if not necessarily actively updated, a series of factors which result in a robust community and documentation system sprawling across gigabytes of code packaged conveniently for inclusion, occasionally clashing but still harmonious in its approach.  Furthermore, I can take joy in the benefits of LaTeX for artwork, discovering that page resizing using, say, the "geometry" package is a breeze in the winds of time saved and can maintain the overall formatting of my creations, and although this joy isn't what led me to explore LaTeX graphic design in the very beginning, considering I was simply attracted by the challenge of using a new creative format, it has kept me around, the admixture of format independence and precise, algorithmically manipulated measurements (not to mention that PGF/TikZ can clipping-mask layers at multiple levels somehow!) has captivated me, fully maintained my sense of awe with plentiful scintillations every single time I accomplish something new in my world of design, because referencing current page nodes, beginning and ending scope environments, and having a text in particular to reference when I have finished composing a piece is frankly incredible, in particular when I begin to consider that the same system handles both my (always properly aligned) interlinear glosses for my undergraduate linguistics work and my abstracted artistic endeavors, a counterexample to both Word's tragically now-widespread horror-story idiosyncrasies - lamentably generalizable to Docs, Writer, any competitor one can think of which has attempted to imitate the shrouded what-you-see-is-what-you-get model - and the difficulties I have experienced attempting to edit Bezier curves by hand in Illustrator or Fontlab/Fontforge, the distinctions between the latter two having blurred away in my mind, and it all constitutes a simultaneous success story and stable, harmonious existence which I have chosen to maintain through the software I use in my day-to-day life.

    Even for the millennia prior to TeX and soon enough LaTeX, history has demanded to progress without cessation, with no stagnation, and with constant documentation of even the most minimal now that there is a means by which we chronologize our being as it runs its course and the inevitability of time threatens to destroy at its own pleasure all which precedes, which furthermore motivates our decisions we make with technology, for the present kind of technological forward motion - technovelocity - encourages faster and faster production with new means constantly morphing between one another or into mutant forms unrecognizable in their design and thus in time ignoring existing systems for innovation's sake, new architectures from the ground-up creating competing standards with the intent of unifying all existing ones under a new, arbitrary definition to be practiced alongside everything which existed prior, thus executing a cycle which I do not at all suggest we strive to prevent, as its occurrence is so frequent as to resemble a nature, though its intensity is more than just certainly a productivity-oriented society's borne fruit which would be uncomplicated and less torn were we to make compromises and overwork ourselves less as we attempt to make wrought the various processes of composition, but even so, we must try not to ignore a fundamental message that the old and often ostensibly illiterate may be the wiser ones risen above the rest with distinction after all.  That is, the arcane system may be better-documented, more explicitly open to the public for examination and modification, more capable or flexible in its architecture or architectural sensibilities, and ultimately more sensical when the careful user compares it to unreadable systems despite complaints of idiosyncratic behaviors and learning curves themselves nullified by a careful, thorough examination of the concept in and of itself in relation to its implementation, even if the explorer is no programmer by any conventional reckoning (I'm certainly not), meaning visual immediacy may perhaps distract from our desire to compose and theorize alongside seeing results with every digital brushstroke where the need to do so may not truly exist in a moment or for the sake of something or another, and we can thus save our thinking and resources for the discrete and raster and, with a sigh of relief that they are in safe, well-composed LaTeX hands having in at least some capacity had their content separated from its presentation, allow our vectors to run free.