Monday, September 7, 2009

Information-as-thing thing...

“Language is as it is used,” could well be the mantra of transformational grammarians. The recognition of information-as-thing form the transformational school is an argument won of preponderance. If that's how people are using the term, then adjust the dictionaries, style guides, text books and what have you.

I come from multiple backgrounds that allow me to accept Buckland's notion of information-as-thing, without criticism. (I should admit that I am a transformational grammarian at heart...) To test the notion, I run it through a few of my fields of participation: art, music and storytelling.

As an artist I know that media, texture, shape, color, etc. (which are at the root of thingness) convey information - are information. The material is as critical in making meaning as the subject of the work. In communications parlance, the medium is the message - or massage. As a musician I recognize that an invisible blanket of reverberating sound waves, both physical and ephemeral, is information-as-thing. Its tapestry of sound informs me not only of the presence of different instrumental voices, but the tempo, volume, chordal structure, etc. indicate to me a sense of motion and emotion. The written music is text, but the performed musical text is also information. The same goes for storytelling: the performance is information-as-process, but the sound is the meaningful conveyance of information-as-thing.

One focus of the essay is the defense of information-as-thing rather than the other two states: as-process and as-knowledge. However, it is information-as-knowledge that I find the most difficult to manage because it is rooted in the often shapeless void of interpretation. Interpretation is required to know which of any various input streams are important, or as the essay illustrates, which pieces of evidence are important. Interpretation provides us with the means to make meaning from chaos – and this process is far more unwieldy than information-as-thing. To me, that is where the ambiguity lies: the evidence can sit before us for a long time before we see its value to our case.

Later, in the discussion of events, which is more in line with my comments above about the performing arts, the author lists three ways in which the evidence of events is used: objects (i.e., bloodstains on the carpet), representations of the event itself (photos, newspaper reports, memoirs), and the re-creation of events. Missing here is a fourth way known widely in archival circles – oral histories, the telling of stories about an event. Buckland regards 'accounts' as “no more than hearsay evidence,” which effectively eliminates the entire world of oral culture from the halls of information science. And yet, later in the essay, he posits the importance of representations in information retrieval. Recollected stories are representations.

I am reminded of other ways oral culture gets shuffled to the side in discussions of information science. Buckland, citing Clarke and Eco, writes that “'Natural sign' is the long-established technical term in philosophy and semiotics for things that are informative but without communicative intent.” Natural sign would be the classification given to animal tracks on the forest floor, scratch marks on tree trunks, cloud patterns, and so on. The Western mind (if I may use large brush strokes) sees no causality in these 'natural signs' perhaps than to ascribe some vague Mother Nature-ishness to them, whereas many indigenous peoples would read these as stories left for them to read and interpret. For example, Buckland sees trees as informative in only two ways: “Obviously, as representative trees they are informative about trees. Less obviously, differences in the thickness of tree rings are caused by, and so are evidence of, variations in the weather.” He sees trees as a scientist or pragmatist does and not as someone whose life or livelihood depends on being able to understand other meanings trees convey.

Parting shot: I felt that the section titled “Being information is situational,” was needlessly complicated. It would suffice to say that the degree to which something is particularly informative depends upon its context – which almost needs no stating at all.

Mark Oppenneer

UIGrrl Responds ... Essays for September 9 class

History of HCI
Sears and Jacko’s historical overview of development of HCI was profound to me, in part as an illustration of how the field’s development mirrored our readings, showing how various interpretations can dramatically morph a culture of sorts, in a relatively short time. The formation of HCI type groups and subsequent terms and language became aligned based on the background of the individuals and on the current situational problems each discipline encounters, relevant to Heidegger’s vision of interpretation of language [Heidegger, Being in Time, 1962]. The history illuminated how powerful words and language are in influencing thought and how we always bring new biases to the picture; truly HCI’s development shows “language is action” [Winograd, Understanding Computers and Cognition: A New Foundation for Design, 1987]. We are creating our HCI world on the fly, in the interconnectedness and sharing of our interpretations.

As a general life principle, as a songwriter I realized the power of language really made me reconsider my own responsibilities when putting out my own ideas and thoughts to others in the world. We all contribute to the twists and turns of culture. For HCI development especially, language may be in fact be the most powerful factor in enhancing (or diminishing) the existence of ‘traditional’ forms of human factors thinking and in fact, our very careers.

On a lighter note, understanding the history and oftentimes disconnection between groups made me feel a little less crazy in my everyday life experiences with UX concerns in a large corporation. With all variants in UX/UI descriptions, roles and responsibilities, sometimes as a group we may appear ununified to the corporate world and individuals out of touch as the field changes so rapidly. But I now can see that much of the confusion is less our own community and more a product of the variables, perspectives and expectations among IS, MIS and companies with various identities, calling upon HCI discipline to solve for different problems. So we’re not just creative crazies! :)

In terms of its maturity, I differ with Sears in that I don’t consider HCI in its infancy or even through an adult stage. It seems more the state of HCI is more like a young-at-heart grandfather – still seemingly ‘young’ and vital due to the input and conversation – but still, at the end of its ‘first’ incarnation as more efficient technology approaches, technology that can better approximate our user needs without such complex mental models and affordances on a screen. We will not diminish the need to broaden our understanding and research around how we think and act, but I think this focus on the interface as we know it will disappear rapidly. I maintain we are on the cusp of something very different in our approach to interfacing with a machine. I provided some comments on this at the end of this essay.

I would agree with the suggestion that information science is the closest and correct affiliation with HCI, because it deals with the process of interpretation and meaning, the most important aspects of human communication and exchange, implicating interaction.

Of all the readings, I have to admit I understood this article least, but did notice some interesting parallels with the other readings. Buckland reviews how information can be thought of, and I believe he purports that in the processing information we begin to regard information-as-thing and when we begin to recognize structured objects in our world present-at-hand to be dealt with, presumably when not in our primordial state of unconscious every day living [Heidegger].

Of particular interest that was similar was Buckland’s assertion that information-as-process is “situational.” This seems to align with Heidegger’s contention in part – that interpretation of information, or in its processing, we include both the original knowledge and the current situation to find meaning.

Also, information-as-thing seems to be related to recognition of objects and properties. It struck me that this type of entity is how computers access information, but people on the other hand, may not naturally think about objects as information, rather they simply process the information and use it passively ready-at-hand in interpreting information. Perhaps these differences may help us understand how to compare and contrast ways we can bridge human and machine intelligence…

I look forward to more discussion on the depth of meaning of this article, especially considering how relevant the term ‘information’ and the discipline of information science (IS) are to HCI today.

Understanding and Being
Winograd essentially provides an overview of the theories of what constitutes understanding and being, drawing from theories of Heidegger and Gadamer [Gadamer, Truth and Method, 1975] as a primer to how people think and therefore how to design for them. He starts with a review of Hermeneutics; foundation for examining the role interpretation should play in the meaning of text. Questions are raised: Is there some absolute truth in a text itself, that we need to uncover? Should we try to rise above our own contexts, to try to understand the original meaning, or is that a fruitless attempt?

These questions of objective vs. subjective interpretation arise also around the current set of assumptions about cognition that persist in our culture today - a ‘rationalization orientation’ that dates back to Plato and Aristotle. Essentially a dual view of reality assumes there are both objective truths in the world and at the same time a reality perceived only through own subjectivity. It states that somehow we can change the physical by our thoughts and intentions in the subjective (just look at the assumptions and popularity of cognitive therapy). (I myself feel biased in believing in this system; having witnessed seemingly magical situations ‘created’ through a focus on words and thoughts).

While debate even now rages about this current thought and about what is true reality – the objective world of objects and things translated, or the subjective experience - Heidegger maintains neither the objective or subject view forms the true understanding of the world. There is a more “fundamental unity” which embodies the notion that that the interpreted (objective meaning) and the interpreter (with subjective perceptions) are both important contributors to meaning. That it’s impossible to understand existence without both.

Further, Gadamer saw the combination of both our current context or situation and our inescapable past (our ‘horizon’) and all the prejudices that come with it. Gadamer maintains we should not be so bent on trying to ‘get over’ our prejudices, because human understanding necessarily draws on both, but remembering that equal consideration must be given to the current situation as it definitely plays its part in interpretation. Therefore, there is always a new interpretation, if only giving way perhaps a few words or actions at a time.

[Note: This would definitely upset many fundamentalist religious people in the world, it would seem – as may feel that the context of the original understanding in the text of the Bible for example is ‘right’ one without regard to current situations and interpretations.]

One more concept discussed is that human awareness and existence is fundamentally ‘primordial’ in that we think instinctively rather than reflectively, i.e., taking time to contemplate our surroundings, something most of us just do not do. This has implications discussed later for anything in which humans act, react or interact with, including design. As humans, we don’t think about the actions we take generally take when performing tasks. And we don’t think about the objects we are using. We are not necessarily deriving meaning or aware of the objects we are using (ready-at-hand); we are just being and doing.

Heidegger continues that when we are “thrown” into situations that interrupt our natural flow we must act, and typically we are suddenly aware of the objects and structure around us that are present-at-hand. Heidegger maintains that therein the state of “throwness” lays the opportunity to take action and produce some meaningful ‘resonance’ in the world – at this intersection of instinct and thought. Winograd later refers to this as the place for innovation in design.

I think though, we have to be careful here – while the space between detached contemplation and instinctual reaction may have meaning, we have to be clear that if we all acted without some real contemplation driving our actions I don’t think we would grow as a culture and society – but that’s for a different essay. 

Direction for Design
In this chapter, Winograd basically asserts that thoughtful design is predicated on understanding the person’s everyday patterns, or sometimes called literal or abstract “conversations” - and discover the objects they interact with however unconsciously. Once determined, designers can design these representations; when the user is then “thrown” into a break down where the flow is interrupted, they may experience familiarity with these objects as part of their work stream, to inform their next move. Winograd says in this space designers have the opportunity to bring forth affordances, shift our patterns of thought, and perhaps engage in new conversations.

Regarding the subject of throwness, HCI scholars may want to investigate it through studying arts and artists. The state of throwness is not just a side effect of living of most artists, but in fact, is the often the desired state of being (I want to get in the “zone”). Abstract painters and jazz musicians, for example, actually utilize and count on being ‘thrown’ ideas – from outside their normal existence. This is the space is which rich innovation lives for them. It is in the combination of the ‘not-thinking’ coupled with some new situation or introduction of new sound, within and recognizing a familiar structure (like a shape on canvas or a repeating chord progression) upon which they act spontaneously. This is the space that produces that rich new sound or an image filled with new colors and light. Also, related to Winograd’s assertions, this is also the space where new, unspoken conversations arise, in the abstract in this case, either between the listener and the player or between the player and his own internal sensibilities, that evokes new thinking/feeling and the unspoken universal “language” of art.

Further, because artists are “experts” at being thrown, they give us the insight, tools and sensibilities with which to adapt and proceed as a society, into the future.

On research and design – UIGrrl shouts from the peanut gallery
Winograd’s assertions on design in practical terms boils down to designers studying users and providing familiar affordances when they get to the state of break down, so they can recover and be on their way again. This all makes sense, but even if we get domains right, or find the correct representation to ‘unconceal’ so to speak, I would not say this is where we should put our focus as future researchers. Granted, as designers there is that space where we can innovate and continue to innovate but let’s remember - we’re not here to find *reasons* to develop our design chops, rather we’re here to make it easier to utilize computing. And to that end, there are bigger fish to fry in the HCI community.

I say we are entering a new era for HCI. The next generation of HCI studies should focus on two important actions:

1) Continued and improved study of users as individuals to uncover their primary conversations. More attention can be paid to methods such as Contextual Inquiry, where we can focus more on these processes and actively try to reduce the need for affordances we think required for users to move on when they break down.

2) Increased study in cooperation with technologists, prototyping networked ‘intelligence’ that can make appropriate adaptations, rather than relying on affordances on a screen. (This is obviously not a new idea, just my thoughts on where we should put our energies.)

To the latter, we should create human-computer or system associations that can anticipate and automate task flow changes before the thrown state actually happens, since it is optimal to keep the user in the conversation rather than fall out of it. For example, if someone from the shipping department depends on the accounting system to work and something happens, how can the system adapt to this change and inform the department or alter the workflows in ways that are still natural to both sets of users? Perhaps the system has intelligence about users on both ends. Perhaps they even know personal things. A very simple example is that the key person, say a manager who signs orders, might be out sick, the system adapts and adjusts the shipping department workflow. A more complex scenario might be that an entire marketing process relies on stock changes, and the system knows about those changes and more importantly, something about the people engaged, so that it can present the most natural prompts or interface possible; the point is that the system can get more detailed inputs about the person’s current situation in a department or domain and can successfully move the train tracks on a new path, seamlessly.

How do we get there? I believe we put our efforts into research and development that finds ways to provide better human inputs into the computer system, thus creating more “intelligence” or put more clearly, an awareness of the individual in the system. One day the acronym UI might stand for “User integration” as opposed to User Interface (and I’ll be called the Mother of User Integration – ha!).

One vision of this future is wearable input mechanisms, where computers or systems intelligence will seamlessly understand the needs of the human. Computers will then support human beings with natural human language (even though it might process it as ‘information-as-thing’). They will be able to respond and prompt the users in a way they understand, removing the screen interface layer and directly communicating with the person. The systems gain greater understanding of the human scenario and can progressively adapt based on that input. This does not refute Winograd’s statements in that the computer should not think necessarily like humans - but rather evolves its own natural processes to understand human thought and need.

Today we have very crude and static examples such as in Google “suggest,” where users are given feedback based on knowledge of their previous input. But imagine a more dynamic scenario where a system can detect even moods or presence and act accordingly - imagine the 2001 Hal computer that notices personal issues like mood and proceeds: “Dave, I notice you are depressed today. Would you like to listen to some of your favorite music and have me change your meeting with John?”

Chris Hawkins
HCI Masters Student
Theory & Research

Information as Created Knowledge

Buckland’s discussion of information as thing spurred a lot of meta-thinking this week as I considered his philosophical claims and practical applications. While I found some of his advice useful, I found his theoretical argument troublesome. After I examine Buckland’s metaphysical discussion of information as thing, I’ll discuss why some of his practical points are still applicable—within a certain context.

I want to discuss two of Buckland’s definitions of information: as knowledge and as thing, since of the three these two are closely related. (It may be that I don’t fully understand Buckland’s use of information as process, but for the purposes of his article and this response, I don’t see much use for it. This point should certainly be a topic of class discussion.)

Despite Buckland’s arguments, I’m not convinced that information as thing can be anything but a metaphor. When he discusses the bits and bytes of a computer, he’s not referring to information qua information but the vehicle through which information is delivered. To think of information as thing is to separate information from the author and the interpreter. It’s to say that information resides in the physical despite humans and despite the process of reasoning. The individual letters in Buckland’s article, for example, are not information per se. It’s the combination of letters into specific words and words into specific sentences and sentences into logical order that present the information Buckland intended.

Contrarily, to think of information as ethereal knowledge in the mind—that is, completely separated from the physical—is to sever the only means humans possess to gather information, their perceptual faculties, from reality. Perhaps someone might argue that illogical whims randomly developing in our minds should be considered information, and they are separated from reality. Let me answer only with an axiomatic claim, the proof of which is self-evident but, ironically, could require more explanation than I have space: All instances of information as knowledge originate from an objective reality that is knowable through the senses. (For an in-depth discussion of this claim, see David Kelley’s The Evidence of the Senses: A Realist Theory of Perception.)

It seems to me, then, that information is factual knowledge formed by an application of reason to existence—a combination of the mind and perceived reality. In other words, I prefer to think of information as created knowledge.

To further point out how information as thing is a metaphor, I submit these three sentences:

“Inside this document is information that will bring down the government.”
“Inside this document is evidence that will bring down the government.”
“Inside this document is dirt that will bring down the government.”

Obviously, the use of the metaphor doesn’t constitute a new definition of information—though I’d be interested to read definitions for information as dirt. Nonetheless, though I presented my problems with Buckland’s metaphysical and epistemological arguments for information as thing, I still find the metaphor useful if we understand it within context.

His discussion of types and tokens, for instance, is an important part of the copyright and intellectual property debate. In the digital era, it’s important to ask, “What constitutes legally protected digital information as thing (metaphorically) and what is digital knowledge?” For example, is a digital song like the Heimlich maneuver, which could not be legally protected because it was knowledge, or is it like sheet music? It is only through the metaphor of information as thing that copyright and intellectual property can exist at all. Even though the physical object, whether book or byte, is not the information per se, authors still hold exclusive rights to copy that information in physical form precisely because their minds created it. Information as thing is the metaphor through which individual rights are linked to property rights but only if the idea of information as created knowledge is properly understood.

Common Ground

Common Ground
What is information? Information as Thing

It all begins with that kernel of “information”, which at first glance sounds so factual and dry. Looking at information more closely, as Buckland and Winograd have, it becomes obvious that information is not so clear cut. We change information from its “knowledge” state to its “thing” state every time we share that information. Convert it into “knowledge” by expressing information in any form and it becomes subjective and personal.

As I read Buckland’s definitions, I wondered that I hadn’t realized that before. It would have helped me have a better understanding of a summer full of healthcare bombardment. I shouted a loud “Amen!” when I read Buckland’s observation that “Sometimes information increases uncertainty”. It seems that both sides of the argument are often saying the same thing, and at other times both sides are clearly stretching the truth (as Buckland says, “Even if we dismiss the argument that untrue information is not information, we could still ask what could not be information?”). But with Buckland’s deeper understanding of information-as-knowledge, and information-as-thing, is it possible for all of our citizens’ healthcare goals to be presented in a way palatable and understandable to all sides? Is there common ground in every argument?

If ubiquitous Internet access has increased nearly everyone’s access to information by an exponential factor, in some situations it has also increased our uncertainty. I think about my own family’s health issues. When I had a fever or sore throat decades ago, I took my doctor’s advice as law, almost sacred. Now, when my daughter has a sore throat and the doctor writes a prescription, I rush home to check the symptoms online, and indulge in long hours of second-guessing. I pay particular attention to websites that meet certain criteria; Does the website I’m consulting for medical advice end in .com (obviously biased to sell a product) or .org (altruistic)?; Does the interface look professional and official (therefore potentially more reliable and knowledgeable than my chosen healthcare provider)? Has the information been presented in such a way to be understandable and comforting, or does it look like it was cut and pasted from the FDA warning in the medicine’s packaging?

How does our interpretation influence information? Understanding and Being

My take from this chapter is that information on its own is not good or bad, but hermeneutics give it the spin that colors the information to be interpreted as one or the other. Winograd talks about objective hermeneutics, or text that exists independently of any interpretation. Buckland might assert that any processing of information would negate any objectivity. Even computer processing has a human being behind it somewhere.

If the town hall meetings I attended this summer didn’t enhance my understanding of healthcare reform, at least they gave me first-hand experience with Heidegger’s “thrownness”. I attended an NAACP meeting in August where the guest speaker was our Congressman. When over a hundred healthcare reform protestors turned up at the meeting usually attended by a dozen friendly faces, the moderator was unexpectedly thrown into a situation where she didn’t have time to think before she acted – she had to react to the events as they occurred. There was no experience or model for the moderator to rely on; if she could have anticipated this overwhelming response, she might have “designed” the meeting so that she could have more control.

How did we get to this current stage of processing information?
A moving target: The evolution of HCI

As I read about the evolution of HCI, I marveled at the imagination and foresight of these HCI pioneers. In a time when only computer operators could physically interact with computer hardware via key punched cards, when even programmers flowcharted and coded on paper forms with little or no contact with hardware, visionaries like Licklider, Engelbart and Nelson were anticipating a future when computers would “become congenial tools that people would choose to use interactively” (Sears, p 5).

And it wasn’t just communication between the human and the computer that was difficult. A whole field of research focused on the difficulties managers had in communicating with people knowledgeable about computers. (Sears, p 7). I have to confess I sort of miss those days – I liked the feeling of superiority my understanding of multi-dimensional arrays and Assembler programming language gave me. Who cares now that I can interpret an MVS core dump?

Looking Ahead
Using computers: A direction for design

Winograd talks about developing aids to bridge the gap caused by breakdowns in human understanding. He mentions training as one of these aids (Winograd, p. 165). But is training an outdated notion? Do today’s users have the patience? We’ve come to expect things to be immediately useful. If a device requires training, we throw it away, replace it. Maybe I just need to adjust my understanding of what training is.

Looking ahead, what will be the next mandatory technology? As Sears says,“software that was discretionary yesterday is indispensable today” (Sears, p. 19). The reverse is also true. Few people have typewriters in their homes anymore, and even landline telephones seem to be phasing out. How many people do you know who don’t have a cell phone? Even Grandpa has an email address. As Sears suggests, perhaps “HCI will become invisible through omnipresence” (Sears, p. 19).

As I’m writing this, my 12 year old daughter is home on her own, and texting me for cooking directions. What is interesting to me is that she is using very few words. She wants to know if the plastic wrap we have it microwave safe, so she texts only “microwavable?” and a photo she has just taken of the roll of cling film. I ask (in full sentences, of course, with completely spelled words and correct punctuation, because I’m a throwback to an earlier age) what she is cooking, and she responds with a photo of a can of Campbell’s soup. This is very different from the conversation I would have had with my mother on the same subject, and illustrates Winograd’s observation that technology is “a vehicle for the transformation of tradition” (Winograd, p. 179). Transformation will happen regardless of our will. We can’t even choose what that transformation will be, because individuals cannot determine the course of a tradition. The actions of millions generate the changes, and the evolution continues.

Liz Foster