THIS IS NOT A STRETCHTEXT
What is a StretchText?
Ted Nelson coined the term StretchText in 1967 as a possible hypertext tool/type. While only a minor concept in the grand scheme of his infamous Project Xanadu, I believe it is an idea that is weighted with a large amount of potential for exploration into the limits of textual and hypertextual meaning making, and understanding.
Stretchtexts are essentially fluidic interactive documents in which users control the quantity of information displayed at any given time. The original vision for this interface was a type of Vanavar Bush style Memex machine with a multidirectional joystick that allowed control over movement through a given work, and control over amount of expansion or reduction of information. One would theoretically start with a relatively simple passage, a summarized version of a longer work. This passage would cover a broader set of data that could inform the user without a lot of detail. When the user wanted more detail, toggling the joystick provided material previously omitted. Subsequent iterations of the piece would become longer and more technical, filling out the user’s knowledge base as they progressed through the text. Thus the text is essentially stretched, as its name suggests. Taking a single piece of information and extending it beyond its initial limits, covering more ground and providing a larger surface area for learning to be detailed and deepened. Like a piece of chewing gum that can be consumed as a whole piece, or stretched thin and potentially extend until its very atoms were linked end on end, touching a practically infinite amount of subjects.The idea is that an expandable page of text provides a seamless flow of uninterrupted learning and thinking. Ideas are evidently conjoined with one another by the relation of the single expandable hypertextual document. Navigation thus becomes less about searching for specific information, and more about seeking a certain level of information.
The Problem In Defining StretchTexts
I believe we find several problems arise when we actually try to design a StretchText system. It occurs to me, given the definitions given by Ted Nelson, and the multiple interpretations that can be found around the web, what defines a StretchText seems to be as nebulous as its interaction is meant to be: boundless and fluid.
StretchTexts are set apart by Nelson and others in their lack of node and link architecture that defines almost all other hypertexts. Other hypertexts are formed out of infinite texts that are interconnected by coordinated links. In essence, hypertexts are externally infinitesimal, while StretchTexts are internally infinitesimal. Hypertexts are interconnected. StretchTexts are intraconnected.
However, in some ways, Nelson himself provides some complications to this simple construction. In the same section of Computer Lib/Dream Machines Nelson also explores the potential of hypertext constructions to be applied to other areas of study. He provides Hypermaps and Hypergrams as examples of hyperlinked interactive media. Yet in both of these tools, there is also a sense of expanding a given field of information. With Hypermaps there is a user-centered control of definition. One can consider the entire globe, or constrict the viewing window to a single street view, all while toggling the amount of information about the various items in-between. Are these not, in a sense, also StretchTexts? But how does that change the meaning of a Stretchtext (let alone the our meaning of text) if visual elements are also being manipulated in the same way? Could we have a video Stretchtext? If a designer created a visual artifact in which there was a compact and unabridged version, would this be an example of a StretchText?
We ask, not for the sake of semantics, but for the sake of fully fleshing out the potential of this concept as a utility. Already, in perusing examples of StretchTexts, there seems to be a level of dissimilitude between Nelson’s original conception and its application in the public sphere. For Nelson the StretchText was a document that grew or shrunk at the basic sentence level. Clauses, phrases, words were added within each iteration of the document the user toggled through. These would add depth, and certainly had the potential for additional clarification or further learning, but this is not initially expressed by Nelson. Contemporary designers of StretchTexts have made this less of a holistic process. Instead, most StretchTexts are designed with individually expandable pieces of information, made evident by hyperlinked text or some kind of button icon, similar to a dropdown menu. The designers of these sites argue that they have created StretchTexts that allow users to more coherently attend to single artifacts and their connected pieces of information, connections that don’t force users to leave or fracture their attention between multiple texts. However, this type of embedment or insertion of information at key points seems to be a similar amount of fracturing. The attention of the user is just divided between a text within the text, rather than a separate text. The interconnectedness, perhaps, provides a different type of understanding that bridges the gap better than a hypertext link, but on the surface does not seem as far removed as StretchText designers would have users believe.
In this regard, are all hypertexts really Stretchtexts?
And, if were honest, to ask this question is really to interrogate how we define the page.
StretchTexts are set apart by Nelson and others in their lack of node and link architecture that defines almost all other hypertexts. Other hypertexts are formed out of infinite texts that are interconnected by coordinated links. In essence, hypertexts are externally infinitesimal, while StretchTexts are internally infinitesimal. Hypertexts are interconnected. StretchTexts are intraconnected.
However, in some ways, Nelson himself provides some complications to this simple construction. In the same section of Computer Lib/Dream Machines Nelson also explores the potential of hypertext constructions to be applied to other areas of study. He provides Hypermaps and Hypergrams as examples of hyperlinked interactive media. Yet in both of these tools, there is also a sense of expanding a given field of information. With Hypermaps there is a user-centered control of definition. One can consider the entire globe, or constrict the viewing window to a single street view, all while toggling the amount of information about the various items in-between. Are these not, in a sense, also StretchTexts? But how does that change the meaning of a Stretchtext (let alone the our meaning of text) if visual elements are also being manipulated in the same way? Could we have a video Stretchtext? If a designer created a visual artifact in which there was a compact and unabridged version, would this be an example of a StretchText?
We ask, not for the sake of semantics, but for the sake of fully fleshing out the potential of this concept as a utility. Already, in perusing examples of StretchTexts, there seems to be a level of dissimilitude between Nelson’s original conception and its application in the public sphere. For Nelson the StretchText was a document that grew or shrunk at the basic sentence level. Clauses, phrases, words were added within each iteration of the document the user toggled through. These would add depth, and certainly had the potential for additional clarification or further learning, but this is not initially expressed by Nelson. Contemporary designers of StretchTexts have made this less of a holistic process. Instead, most StretchTexts are designed with individually expandable pieces of information, made evident by hyperlinked text or some kind of button icon, similar to a dropdown menu. The designers of these sites argue that they have created StretchTexts that allow users to more coherently attend to single artifacts and their connected pieces of information, connections that don’t force users to leave or fracture their attention between multiple texts. However, this type of embedment or insertion of information at key points seems to be a similar amount of fracturing. The attention of the user is just divided between a text within the text, rather than a separate text. The interconnectedness, perhaps, provides a different type of understanding that bridges the gap better than a hypertext link, but on the surface does not seem as far removed as StretchText designers would have users believe.
In this regard, are all hypertexts really Stretchtexts?
And, if were honest, to ask this question is really to interrogate how we define the page.
How Do StretchTexts Mean?
For Nelson StretchTexts are a kind of ThinkerToy, an application that in the tradition of Marshall McLuhan, creates a different level of meaning by its means of interaction. The simple version of current hyperlinks on the internet are without structure, and mostly interruptive and self-promotional. The majority of the links presented to current internet users are advertisements and other unrelated materials, similar-to lists, and suggestions-based-on-history. This is of course because designers do not perceive the medium as making meaning, it is only a source of potential revenue, banking on distraction and fractured disassociation.
This creates a kind of spotty linear directional thinking. Users, theoretically, begin to see each distinct artifact as minimally valuable in relation to the wide variety available. Clicking links has been called site “hopping” or “jumping”, perhaps affirming for us this societal sense of nomadic usage. Links become means of escape, opportunities for completely new information. Infinite novelty, in other words, salience that is completely determined on the user’s ability to disengage as easily as they engaged with the material. The interaction itself, arriving in order to leave, attending in order to remain ignorant, may become an end unto itself. The activity is understood, but the content is lost.
A structured hypertext, on the other hand, has the potential to mean something more, and create new meaning that is dependent upon the user’s earlier understanding. Structure provides a sense of connection and position. Meaning is created by an understanding of multidirectional movement in a tangible sphere of information. Those at Project Xanadu have even made a three-dimensional program to make this more self-evident. Links no longer create a sense of distance, but closeness. One piece of information is brought into immediate relation with another piece, thus allowing for new and perhaps unexpected connections. I consider this in light of the work Halavais conducted on the links in a number of blog posts, finding that a majority of connections were limited to a few major cities, thus, in effect, creating a large amount of understanding without reference to the ideas and perspectives of the majority of people. Hypertextual structures provide a multivalent understanding that evolves as designers share equally in the task of interconnection for the sake of fuller understanding.This creates a kind of spotty linear directional thinking. Users, theoretically, begin to see each distinct artifact as minimally valuable in relation to the wide variety available. Clicking links has been called site “hopping” or “jumping”, perhaps affirming for us this societal sense of nomadic usage. Links become means of escape, opportunities for completely new information. Infinite novelty, in other words, salience that is completely determined on the user’s ability to disengage as easily as they engaged with the material. The interaction itself, arriving in order to leave, attending in order to remain ignorant, may become an end unto itself. The activity is understood, but the content is lost.
So what does it mean when we stretch a text? Our sense of movement changes, we are no longer jumping, nor are we locating ourselves in relation to other material. The user retains a sense of non-motion while the artifact itself shifts character. There is a pulling open, a revealing nature that is implicit within the action of stretching. While hypertexts imply that the user can make new connections between items of information, StretchTexts imply that users are in control of a level of ignorance that the mode provides. To rip open a StretchText is to rip open the perceived limits of one’s knowledge. Hypertexts create new perspectives every time they are interacted with, while Stretchtexts render a more defined understanding post-interaction. Whether this definition is a solidifying of quality or quantity is hard to pin down. One may argue for either. Certainly quantity of information changes, additional information results in a larger amount. The debate perhaps resides in whether more is always or necessarily better on principle. Does adding more deadwood to a sentence make it more defined? Does it mean differently for the added nuance? This is subjective, and perhaps too broad a question to ask, and must be determined on an individual artifact basis.
However, the true distinction between these two modes of meaning-making seems rather arbitrary. When interacting between the artifacts within a hypertext, one is, in a sense, stretching the meaning of the original artifact, the connections that are being made add depth and help define the initial understanding. Is there, after all, any real difference between inserting an entire article into the middle of a sentence via a dropdown menu and simply providing a link to the same article? Do they not both provide a sense of hierarchical meaning making and connection that embed a deeper level of knowledge within a shallower? Perhaps what we are really asking is whether these seemingly separate conventions are really just artificial distinctions that have the same result? While I wonder, on a theoretical level, if this might not be true for all hypertext arrangements, in practice, for a general audience, there will be a shift in meaning that is determined by the means of linking.
Readability – StretchTexts vs The Book
Paula La Forge indicates to us that digital media has the same potential as print media for cognitive load on the user. That, in fact, because of the multifaceted and variable nature of ThinkerToys, like StretchText, they have the potential for even higher cognitive loads that print, since print is static.
In terms of readability, StretchText documents differ very little from other digital sources, the discussion of which Marshall covers quite nicely, indicating that, from an objective view, there is very little difference as long as certain conventions of reading in general are maintained and attended to. As a text document, StretchTexts can easily be manipulated to fit any given set of print conventions: spacing, paragraph length, font etc. StretchTexts, as malleable entities, have the added advantage of principally being in the users control, thus allowing designers to potentially build in features that might changes some of these settings to the preference of the user. And again we may be reshaping the bounds of this medium, is an in-page setting to change font size perhaps a form of StretchText? Is the user changing, perhaps ever so infinitesimally, the individual meaning of the text for themselves by such an act? Marshall would probably say: yes, indeed it is!
StretchText, in a way, resembles Drucker’s questions of the conventions of reading, however. The text is permeable, and therefore unstable. A user might be given the sense of a deepening or refining of knowledge. On the other hand they may also be given a sense of disillusion or deception by the artifact. Perhaps what they wanted to know was hidden farther down the chain of stretches than they were willing to follow it. So while StretchText may seem like a boundless medium that frees the user from the structure of print, it may still necessitate an awareness of user expectations, and a provision of utilities and conventions to make navigation less daunting or obscure. A careful choice must be made, on the part of the designer, in crafting the separate version of a document, and deciding which parts should be optionally expandable, but are likewise not necessary for a complete understanding of the original document.
As I have noted elsewhere, Bernstein comments that navigation in hypertexts are something of a nonissue. That most users have some kind of intuition about their digital orientation in a hypertext, and those who don’t are liable to be surprised by some kind of new learning anyway, as he also indicates that perhaps centralized navigation is not as helpful as Krug might infer. In terms of text navigation, some might find StretchText to be a superior medium, in some regards. Its use of embedment is of course not entirely original. Print texts have long had footnotes that navigate in a similar fashion. There is an indication that there is an authorial elaboration, using the asterisk symbol, one is linked to that further discussion, not altogether different from a drop-down menu of additional information. This is not, however, as fluid as a StretchText, and the physical limits of a page would, conventionally restrict these kinds of stretches to a few short clarifications, and only a single level of refinement. Navigation in a Stretchtext, though infinite, always has a sense of centrality. No matter how far “down” within the various stretches one delves, there is always a sense of location within an initial meaning, the connection is self-evident. One has only to backtrack to reaffirm connections, but bearing is never truly lost as one never, theoretically, leaves the artifact, as one might in another hypertext format. Going back, therefore does not involve a search through a web of connections, but retracing by constriction of information. Therefore position within an article is never judged spatially, but qualitatively. The information itself becomes the focus of attention, not the place of the artifact itself. Which may be part of the revolutionary nature of StretchText if applied on a larger scale.
In terms of readability, StretchText documents differ very little from other digital sources, the discussion of which Marshall covers quite nicely, indicating that, from an objective view, there is very little difference as long as certain conventions of reading in general are maintained and attended to. As a text document, StretchTexts can easily be manipulated to fit any given set of print conventions: spacing, paragraph length, font etc. StretchTexts, as malleable entities, have the added advantage of principally being in the users control, thus allowing designers to potentially build in features that might changes some of these settings to the preference of the user. And again we may be reshaping the bounds of this medium, is an in-page setting to change font size perhaps a form of StretchText? Is the user changing, perhaps ever so infinitesimally, the individual meaning of the text for themselves by such an act? Marshall would probably say: yes, indeed it is!
StretchText, in a way, resembles Drucker’s questions of the conventions of reading, however. The text is permeable, and therefore unstable. A user might be given the sense of a deepening or refining of knowledge. On the other hand they may also be given a sense of disillusion or deception by the artifact. Perhaps what they wanted to know was hidden farther down the chain of stretches than they were willing to follow it. So while StretchText may seem like a boundless medium that frees the user from the structure of print, it may still necessitate an awareness of user expectations, and a provision of utilities and conventions to make navigation less daunting or obscure. A careful choice must be made, on the part of the designer, in crafting the separate version of a document, and deciding which parts should be optionally expandable, but are likewise not necessary for a complete understanding of the original document.
As I have noted elsewhere, Bernstein comments that navigation in hypertexts are something of a nonissue. That most users have some kind of intuition about their digital orientation in a hypertext, and those who don’t are liable to be surprised by some kind of new learning anyway, as he also indicates that perhaps centralized navigation is not as helpful as Krug might infer. In terms of text navigation, some might find StretchText to be a superior medium, in some regards. Its use of embedment is of course not entirely original. Print texts have long had footnotes that navigate in a similar fashion. There is an indication that there is an authorial elaboration, using the asterisk symbol, one is linked to that further discussion, not altogether different from a drop-down menu of additional information. This is not, however, as fluid as a StretchText, and the physical limits of a page would, conventionally restrict these kinds of stretches to a few short clarifications, and only a single level of refinement. Navigation in a Stretchtext, though infinite, always has a sense of centrality. No matter how far “down” within the various stretches one delves, there is always a sense of location within an initial meaning, the connection is self-evident. One has only to backtrack to reaffirm connections, but bearing is never truly lost as one never, theoretically, leaves the artifact, as one might in another hypertext format. Going back, therefore does not involve a search through a web of connections, but retracing by constriction of information. Therefore position within an article is never judged spatially, but qualitatively. The information itself becomes the focus of attention, not the place of the artifact itself. Which may be part of the revolutionary nature of StretchText if applied on a larger scale.
Usability – StretchTexts vs Krug
On the whole, the nature of StretchTexts does not necessarily deviate from Krug’s vision of the internet and usage guidelines. StretchTexts, after all, are a way to format documents and arrange levels of information, but does not necessitate the disavowal of other digital conventions. Perhaps the pages the StretchTexts are found on still have permanent navigations, or search utilities, or advertisement banners, or Site IDs and taglines. Krug advocates for a no muss, no fuss digital sphere in which the tools for information gathering are not overly complex. In that regard StretchText may in fact be Krug approved. They are created around the inference that different users desire different densities of information, which they fully control within the medium. The tool for learning is entirely self-evident, clicking the link or drop-down renders clarifications, details, additional examples, or definitions all pertaining to the immediate term, phrase, or idea the user desired to obtain more information about. The conventions, as we have before discussed, are not altogether new, therefore perhaps not requiring any extra effort on the user’s part to acquire them. The embedded nature of StretchText may act as a kind of inherent signposting, as before discussed when reviewing navigation.
However, StretchTexts represent a kind of antithesis to Krug’s ideals. Krug’s primary mantra is to not make users think. This is applied in the same sense that Nelson’s ThinkerToys should only result in complex thought, not make confuse users with its own complexity. For Krug, there seems to be an extension beyond this maxim. That in general, users don’t think anyway. That the designer’s goal is to hide the subtle effects of the medium, so that the subliminal message can work in the favor or powerful designers preying upon hapless users mindlessly clicking and pleasantly surprised at how easy it is to find what you are looking for and purchase it. This may be my own cynicism breaking through the fourth wall here, but I also believe that users will, inevitably, think regardless of Krug’s admonitions.
A nightmare world for Krug might resemble one in which StretchText is the primary medium for information transference. The very nature of how StretchTexts mean might put an end questions of scannibility. StretchTexts encourage the opposite of scanning or browsing, they encourage development and careful consideration. They ask to user to be fully aware of the system with which they interact with. Not obtrusively, but in the way that most hypertexts make each click a meaningful act, rather than the mindless one that Krug sets as fact. Even at its best, Krug’s vision is of an incredibly ignorant user base, numbers and statistics rather than real people with incredible capabilities for new perspective and innovation. Perhaps not a great business model, but a world of StretchTexts might be a better one, or at least a more intentionally educated one. This is part of my problem with hypertext structures as well. They depend too much on the user as infinite wanderer. Even the best built hypertextual internet would be so vast that a lifetime would never suffice to give the user what they might need. Too much fractured attention, even if well connected, maybe not enough of Krug’s self-evidency there. StretchTexts provide a worthy alternative to these polarized digital choices. Users maintain a stricter control, the connections may be more localized, but they offer a more specific depth that will prevent nomadic scanning practices.
However, StretchTexts represent a kind of antithesis to Krug’s ideals. Krug’s primary mantra is to not make users think. This is applied in the same sense that Nelson’s ThinkerToys should only result in complex thought, not make confuse users with its own complexity. For Krug, there seems to be an extension beyond this maxim. That in general, users don’t think anyway. That the designer’s goal is to hide the subtle effects of the medium, so that the subliminal message can work in the favor or powerful designers preying upon hapless users mindlessly clicking and pleasantly surprised at how easy it is to find what you are looking for and purchase it. This may be my own cynicism breaking through the fourth wall here, but I also believe that users will, inevitably, think regardless of Krug’s admonitions.
A nightmare world for Krug might resemble one in which StretchText is the primary medium for information transference. The very nature of how StretchTexts mean might put an end questions of scannibility. StretchTexts encourage the opposite of scanning or browsing, they encourage development and careful consideration. They ask to user to be fully aware of the system with which they interact with. Not obtrusively, but in the way that most hypertexts make each click a meaningful act, rather than the mindless one that Krug sets as fact. Even at its best, Krug’s vision is of an incredibly ignorant user base, numbers and statistics rather than real people with incredible capabilities for new perspective and innovation. Perhaps not a great business model, but a world of StretchTexts might be a better one, or at least a more intentionally educated one. This is part of my problem with hypertext structures as well. They depend too much on the user as infinite wanderer. Even the best built hypertextual internet would be so vast that a lifetime would never suffice to give the user what they might need. Too much fractured attention, even if well connected, maybe not enough of Krug’s self-evidency there. StretchTexts provide a worthy alternative to these polarized digital choices. Users maintain a stricter control, the connections may be more localized, but they offer a more specific depth that will prevent nomadic scanning practices.
Application – StretchTexts on the Web
How couldn’t we use StretchTexts? Imagine Wikipedia articles that could be expanded into full academic documents and interlinear discussions--forums, critiques, and sources all on one page, embedded within itself. Kindle versions of novels that could freely transition between summarized, abridged, unabridged, and annotated versions of the text. Perhaps a grad student, working on Seminar and Conference Versions of a paper could transition between views of the two within his word processor. Not to mention articles, magazines, and newspapers that could all benefit from this medium.
Fiction is a relatively unexplored realm in which to use StretchText, although there are a plethora of hypertextual narratives that utilize similar formats. StretchText strikes me as more applicable to non-fiction, perhaps more informational sources. But hopefully, I have demonstrated the greater bounds of the medium that bridge the gap between the nebulous fracture of some hypertext narratives, and simple linear storytelling. It might perhaps make hypertextual storytelling more accessible to a wider audience, has more direct connection and meaning making, yet still alows audience to change their experience with an artifact, make new connections, etc.
Despite their apparent usefulness, there are only scant examples, currently, around the web that use this structure, with the intention to create a new StretchText based meaning:
Despite their apparent usefulness, there are only scant examples, currently, around the web that use this structure, with the intention to create a new StretchText based meaning:
The Eclectic Light Company
A blogger by the name of hoakley uses this site to blend two loves. Writing short articles about art, particularly analyzing paintings from the 17th-19th Centuries, and technology. The article I have linked above is from a short series of experiments this blogger made in using StretchText, even providing a tutorial on how to use StorySpace to create one.
Blog Architecture
Another blogger, by the name of Ted Goranson, uses this blog as an experiment in "spatial hypertext". In application, this evinces itself as dropdown text boxes that expand from menus. However, the content offers a seeminly endless supply of nuanced thought on the structure of hypertext and the importance of the structure in reference to meaning-making.
Coding In Paradise
Brad Neuberg, a software engineer, developed a StretchText in JavaScript, which he shares on the above page. He also demonstrates the use of this StretchText, showing off the smooth transition of embedded material into the main text, as well as conventions of salience for what is clickable with dotted line boxes. If I had had the skills, this is the kind of StretchText I would have liked to produce for this project.
Tinderbox StretchText Writing System
Another program for creating StretchTexts, seemingly quite well developed. There are additional conventions added here, as well as sidebar notes that appear on the margin of the page, an addition to Nelson's simpler version of StretchText.
Conclusions
This is not a Stretchtext...Or is it? One of my main concerns in producing this piece was determining whether the mode by which I delivered my analysis adhered to or challenged said mode. I wonder if I have done both and neither. This question, and one that is repeated throughout, is integrally linked to how StretchTexts, and by extension all hypertexts, mean, and make meaning, or shape understanding. I have tried to flesh that out throughout, but at times I do wonder if such an awareness really frees the user or enables them to make new meaning. I wonder if we put too much pressure on the medium, though it certainly is a message unto itself, to make that meaning for us. We can determine the possible change in meaning making without ever having to use the medium we are analyzing or speculating about. This was certainly true for Nelson, whose hypertext structures were concepts before they were reality. Perhaps what I am saying is about as anti-Krug as they it gets. People naturally think, and intentionally make choices about what they desire to perceive. Even with the best tools, people might ignore them, move on to portions of the internet they can continue to scan and click or scroll mindlessly. Antithetically, despite the culture of the internet as it is, there are still people who are thinking deeper, making connections, and developing ways of re-seeing the world using the engines of meaning making they have.
And as I have, I believe, done that here, then perhaps the answer is yes. But that’s also not what really matters.
