- Variability: "A new media object is not something fixed once and for all, but somethingthat can exist in different, potentially infinite versions" (36). Manovich lists seven examples ofvariability common in contemporary new media, and also considers more foundational differencesvariability enables: for example, hypermedia elements and structure need not be "hardwired" as inold media. Variables replace constants, and data separated from algorithms (as in computerprogramming). But to some extent this variability is radically limited to selection from a group ofpre-packaged forms: a concept Manovich will later expand as "selection."
- Transcoding: the "reconceptualization" which occurs during computerization, thetransformation of media into computer data. The mapping of concepts such as plot, sentence, familyportrait, or summer blockbuster into the computer's text, packet, pixel, or other data structure,creates a composite "blend of human and computer meanings" (46). (In computer science, the term"transcoding" itself signifies movement of data between formats.)
The last of the five principles, transcoding, is "the most substantial consequence ofthe computerization of media" (45). Manovich suggests thinking of new media as "two distinct layers the 'cultural layer' and the 'computer layer'" (46), though he proposes this distinction in amanner which does not imply disconnection between the two. "Transcoding" facilitates complexrelationships between the systems of organization of culture and the means by which we affect thosesystems in computing. For example, conventions of computing interfaces influence the design ofhypermedia. Students instructed to "build a database of information" may proceed differently thanthose told to "take notes and organize raw material." Programmers may design computer-basedinterfaces and shape media formats based on cultural objects, like the controls of media players,which emulate the VCR.
In other words, Manovich argues that computer and culture influence each other. Of course, this isnot a groundbreaking assertion, but embedding that relationship into the theory of new media sets atone for the text and distinguishes Manovich's work from others which portray the relationshipdeterministically. Sadly, the dominant common sense of computer design and use follows thedeterministic pattern. As Robert Johnson notes in User-Centered Technology, computers areoften thought of as "black boxes," and the humanagents and cultural influences which shape the design of their interfaces and use patterns areinvisible. Manovich's principle of "transcoding" illustrates the gravity of this misconception.
To further separate these five principles from other conceptual problems, the next section, "WhatNew Media is Not," debunks six commonly accepted assumptions about new media. Once again, Manovichrelies on comparisons between cinema and new media, as he systematically demonstrates that neitherdiscrete representation, random access, or multimedia are the unique province of new media. They areproperties present in cinema as well. Likewise, claims that new media are new because of"digitization" and "interactivity" fail under critical pressure. The periodic sampling oftenconsidered unique to digitization is at the heart of film, and considering any medium "interactive"is a mere tautology (50, 55). But this is not merely a terminological clarification, though itcertainly serves that purpose, as demonstrated below; it is a reminder that "newness" in and ofitself may not correspond to significance.
New Media and CinemaReaders of The Language of New Media may be tempted to misrepresent or simplify therelation of new media and cinema that Manovich carefully develops over the course of thework. Indeed, the cover art a heavily manipulatedphotograph of film stock which literally wraps around the text gives some credence to thenotion that Manovich's argument is simply, "New media works like film." But this sort of judgmenthas obvious flaws.
Some reviewers have configured the sixth (and final) chapter as a "coda" or "envoi," and with goodreason. The fifth chapter ends with a shift to the past tense, a ruminative final paragraph whichbegins, "In this book, I have chosen to emphasize [...]" (285). Manovich admittedly structures the text so that the sixth chapter is a reversal whichreflects back upon the first five (12), a design to some extent represented in the form of thisreview. But he also notes that the chapter continues the trajectory of the book as a whole. And,most importantly, it's possible that Manovich downplays the nature of the reversal which does occurin "What is Cinema?" to achieve greater rhetoricaleffect.
For Manovich cinema provides a double influence: film theory is the "key conceptual lens" (9) withwhich he investigates new media. Film, especially the work of the Russian avant-garde, and DzigaVertov in particular, is the primary source of explanatory examples. "Vertov's dataset," Manovich'scollage-like prologue, is a series of stills from Man with a Movie Camera accompanied byquotes from the text. The stills reappear throughout the work, at section breaks, and in fact arethe only visuals included in the text.
Unfortunately, this focus minimizes the effects of print literacy on new media. Manovich notes someareas where the influence of print is apparent. "Cultural interfaces rely on our familiarity withthe 'page interface,'" he notes. "Given that the history of a page stretches back for thousands ofyears, I think it is unlikely that it will disappear so quickly" (74, 75-76). Manovich calls on thework of Roland Barthes when tracing the genealogy of the screen from Renaissance painting throughprint to cinema (104), and again when arguing that the history of the logic of selection predatesthe development of new media (125). But the influence of the forms of print culture is overshadowedby the power and influence of cinema.
In some ways, this focus is simply a function of remaining true to the established method of arecord of the present, and recognizing the "general trend in modern society toward presenting moreand more information in the form of time-based audiovisual moving image sequences, rather than astext" (78). There are many specific representations of this trend. Writing of the work of virtualreality programmer and theorist Jaron Lanier, Manovich argues that the repression of linguisticforms in new media is a continuation of "the fantasy of objectifying and augmenting consciousness[...] the desire to see in technology a return to the primitive happy age of pre-language,pre-misunderstanding" (59). Shifting to the emergence and influence of new technologies at the turnof the twentieth century, Manovich argues that cinema "impressed itself [more] strongly on publicmemory" than forms of electronic communication which emerged at approximately the same time, because"the ability to communicate over a physical distance in real time did not seem by itself to inspirefundamentally new aesthetic principles the way film or tape recording did" (162). The traditionalstyle of photography and cinematography with its "linear perspective, depth of field effect [...]particular tonal and color range, and motion blur" (179) towered over visual culture, shapingemergent computing technology, yet was seldom foregrounded as a certain kind of realism(191-92).
But though dominance of new media by cinematic forms is a representation of the current status ofthe relation between technology and culture, Manovich shows this need not be the case. Theavant-garde, in particular, have always resisted conventional models. Vertov's dataset derived itslasting strength from that conscious differentiation. Notably, continuation of the tradition ofquestioning established forms occurs when new media designers cope with the technical limitations ofcomputing. For instance, the first versions of the Mac OS were monochromatic, because display technologies simply couldn't representcolor very effectively (63). In its infancy, QuickTime worked best with very small framesizes. These limitations forced an interesting historical convergence: