Research

From CS260Wiki
Jump to: navigation, search

Bjoern's Slides

File:Cs260-slides-09-research.pdf

Extra Materials

Discussant's Slides and Materials

Reading Responses

Kurtis Heimerl - 9/28/2010 14:08:48

The Structure of Scientific Revolutions Oh, I love this book. I read it as a kid, and I think it fundamentally shaped me into the misfit academic I currently am. There's so much here, I'll try to stay on topic.

The biggest thing that jumps out of this work is its analysis of what makes science science. The lack of "revolutions" in mathematics is a strong indicator of the fact that mathematics is not a science. I'd argue most of computer science falls into the same bucket. Has the push to collaborative work come from the fact that HCI couldn't solve collaborative problems with existing techniques or the ever-expanding scope of any scientific endeavor? Undoubtedly, it's probably both. As an example from my own field, distributed systems were built because scaling monolithic ones was becoming harder and harder. Does this fit the book's thesis? I think so, but my bias may just be showing.

One interesting way to look at this problem is by comparing social science, physical science, and liberal arts. The nature of these revolutions is that they explain the data in a fundamentally better way. Because of that, it's very hard to find an 80 year old chemist who doesn't "believe" in much of the modern literature. However, it's quite possible to find this in something like sociology; the issue has never been clearly resolved. Lastly, in Regional Studies, each professor is defined by their belief structure, rigid, unmoving, and determined as early faculty.

I won't even get started on the discussion of where engineering falls in this whole design space.

None of this is fundamentally related to the point you were trying to make with these readings. I think they're designed to show us that we don't have to build the entire building, that science is the mundane part and we should enjoy it. I heartily disagree. I think that much of computer science (and HCI in particular) is being done in the wrong paradigm and the results are of less value because of it. With this belief, I cannot, in good coincidence, do mainline HCI research without feeling like I'm spinning my wheels. This bodes poorly for my chances at a faculty position.

Pasteur's Quadrant 48 pages of reading for a 1-night space? Really?

This paper angered me at a deep, fundamental level. The general nonsense about technology being outsourced to Japan (oh noes, the sky is falling!), the argument that someone categorizing birds is not on a quest for fundamental understanding, and a number of other points, are just wildly off. I don't want to go into too much there though, as there's actual positives to take.

Firstly, I question why you had us read this. Are we to take stock of where our research lies? HCI is an engineering discipline, and therefore unlikely to contribute anything fundamental (at least from a physicist's POV). Also, a great deal of HCI work is not application facing. Does it then fall into the bird-watching bundle?

These distinctions are meaningless to me. I suppose there's some validity, from a policy perspective, of directing your research dollars. At the same time, I can't think of any research with absolutely no potential real use. From my view, the question is simpler: "Are you addressing the hard problem?". Sometimes the hard problem is theoretical (quantum physics for instance) and sometimes it's engineering (most of computer science).

I'm a bit confused about the thesis of my response, actually. I suppose I have a natural push away from bucketing. I'm not sure if that is actually a meaningful contribution to this discussion of this work.



Anand Kulkarni - 9/28/2010 14:55:03

Transforming the Paradigm

The author discusses how to categorize research, in particular focusing on the dynamic between basic and applied research.

The core contribution is a new model for understanding the relationship between basic, applied, and intermediate research, along with a historical overview of the development of models. The author suggests that the understanding given by his new model can be used to guide policy development -- for example, in understanding how research sponsors should support basic or applied research.

The author's argument isn't really supported thoroughly; it's developed through a series of examples, quotes, and discussion. Evaluating such papers can be difficult, since the author doesn't convince us early on of the importance of his problem, but the model seems reasonable. The intermediate case of "use-inspired basic research" seems like an all-encompassing category rather than a tremendously novel or useful one. The relationship isn't always so cleanly defined. This is a "soft" paper, so no data or experiments are really appropriate, but I wish the author had discussed more implications of his work and given more motivation for his problem up front.


Structure of Scientific Revolutions


Kuhn discusses the creation and value of paradigms in the practice of science.

Understanding paradigms can have value not only in HCI but in any field of research, as it lets us understand what standards are required for work to challenge a dominant paradigm, and also why certain modes of thought persist in practice even when they are substandard or in contradiction with pieces of empirical evidence. Characterizing the practice of science as reconciling contradictions with theory helps us understand better the value of other papers we evaluate in HCI or in other fields of research. It can also help us better identify new research opportunities.

Kuhn's argument is heavily developed through a series of major examples of paradigms in the history of science, such as the replacement of Newtonian mechanics with Einsteinian mechanics. This is again a soft argument, but it is supported well with historical information and is a fairly compelling and complete portrayal of how paradigms operate. I wish Kuhn had proposed solutions for better ways to reconcile errors within a paradigm; in his defense, this has been done by many subsequent students of his work.


Airi Lampinen - 9/28/2010 15:09:56

Kuhn's classical "The Structure of Scientific Revolutions" gives an outline of "how science works". In this text, Kuhn outlines his famous vision of how science proceeds not as a neat, continuous accumulation but as a structure within which "normal science" is every now and then interrupted by a paradigm shift that happens when a discovery or new theory revolutionizes the scientific landscape and forces scientist to revisit their starting points and explanatory models.


The biggest question in my mind, while reading Kuhn for a HCI class is how well his model applies to research that does not fall to the realm of natural science. For instance, multiple paradigm "shifts" can be pointed out from the history of social psychology (from behavioral to cognitive to linguistic) but it is common to think that social psychology in its contemporary form is a science with many co-existing paradigms. Hence, a radical shift in thinking does not necessarily completely undermine the previous paradigm but they may, in some cases, exist parallel to one another. I guess this could be explained in a Kuhnian way by stating that social science, then, is not a mature science, but I'm doubtful whether such maturity can ever be reached.


HCI, then again, is intimately related to the advancement of technology, not only theory and empirical research. This means that the object of study is changing in a way that to me seems different from the change occurring in nature. Futhermore, new technologies may address entirely new problems or they may provide solutions that are superior to the older ones. This, however, does not necessarily undermine the correctness of older technological approaches. In a way, paradigms shift, but at the same time, it seems that these revolutions are different from those in natural sciences. Hence, I wonder whether it makes sense to apply Kuhn's model to all fields of research, as is so commonly done.


Stokes' text on "Pasteur's Quadrant" discusses the nature of research from a different perspective, addressing the division of basic and applied research and showing how such a dichotomy is very problematic. Stokes describes different suggestions to overcome the problem and then goes on to present the quadrant model of scientific research, a 2x2 matrix that categorizes research based on whether it is influenced by considerations of use and a quest for fundamental understanding.


The quadrant model of scientific research is a useful way to overcome the dichotomy between basic and applied research and underline the issue that considerations of use and understanding are not necessarily in contradiction. When applied to HCI or research on new technologies even in a wider sense, this remark is not very surprising. However, it seems that the dichotomy of applied and basic research lives strongly in the minds of many and makes it hard to accept that use and understanding can at times go firmly hand in hand. Hence, the quadrant model is a valuable contribution in showcasing this issue with strong examples.


Stokes also describes well how all the quadrants of the matrix have their place and purpose in the whole, even the part that is driven neither by considerations of use, nor by a quest for fundamental understanding. This observation was the biggest discovery for me, as I had earlier seen only interpretations which state that "no one wants to be in that fourth quadrant". It is clear that most researcher definitely (and rightly) want to keep out of it but that there is research that justly falls to that category and still can be valuable for certain purposes, is something that is pointed out very rarely.


Dan Lynch - 9/28/2010 16:13:35

The structure of scientific revolutions

This was a mind-bending article that focuses on the concepts and perspectives of great thinkers and scientists in our past. In particular, what were these people thinking with respect to what was capable at that time? Its easy for us to analyze what we would do different because we have benefited from history, but how can we analyze their revolutions for their time? The notion of conceptualizing ideas and elements of arbitrariness is discussed, in the sense that we as scientists may require some type of stratification in order to proceed with research.

In the section on “Normal Science as Puzzle-Solving”, the author describes how scientists work within a particular sector, or a manner in which scientists work he coins as normal science. He claims that they don’t need to match the laws, just take data: “Coulomb’ measurements need not, perhaps, have fitted an inverse square law”. Now if I interpreted this correctly, then this means that sometimes its good to be blind in some sense while doing research and taking measurements. You may find unexpected, novel results that you may not have found if you were fitting your data to the laws that exist within a certain paradigm.

This is important on a philosophical level in that you sometimes need to step outside of the box and go on with solving a puzzle that you find challenging. You may find something that benefits humanity!

Transforming the Paradigm

This article discuses our views of science, in particular the change over time of these views... Acknowledging the change and fundamental shifts in perspectives can also be done by looking at early dissent, or people in history who thought differently: “once the prevailing paradigm is challenged, it is not difficult to find early observers who tried to reshape its one-dimensional images” (59). The notions of applied versus pure, or basic are mentioned. This is most easily understood when thinking about mathematics. Abstract algebra, for example, is very much a pure math, whereas electrical engineering and Fourier transforms would be an applied math. These two schools of thought define things in very fundamentally contrasting tones, and its important to understand these differences.

One diagram portrays the fundamental differences in that the applied research has an aim, or ultimate goal, whereas the pure research is depicted as stagnant. Now, I think this could be the case, however, I must say that the diagram is flawed. It should look more like a vien diagram, where the pure and applied have some overlap. This is because without the basic fundamentals, the applied researcher would have no basis for which to conduct research. Thus, its important to have an understanding of both to optimize research and scientific progress. The best model in my opinion was the Australian Modification of Linear Model, where strategic research it depicted somewhere in the middle between a continuum of applicable vs. abstract research.


Matthew Chan - 9/28/2010 16:46:54

===Pasteur's Quadrant: Transforming the Paradigm==

This is by far one of the densest materials i've ever read--probably because it's more historical and reading about "basic" research vs. "fundamental" vs "purposive" vs. "mission-oriented" and several degrees of research and what constitutes into each category. However, this paper does provide a unique insight into the lives of scientists right after World War II. Namely, science and the creation of the atomic bomb help seal victory for the Allies, radar had a place, etc. so it was only right for science to continue advancing. In Vannevar Bush's paradigm about basic science and its role in technological innovation, many dissenters materialized and started creating different forms of the paradigm since many research just didn't fit, ie. applied vs. basic research and whether to seek understanding or an aim. I think this is a fairly important document since it highlights the historical aspect of science and research today, especially the National Science Foundation and how some research gets funding and others don't. There were no methodologies or techniques, but lots of exploration of different historical figures such as James Conant, Gerald Holton, James Irvine, Harvey Brooks, and many more and their ideas of research. This paper doesn't relate to my field of work, but i can understand how it might have affected my research/work to be what it is today. Ultimately, Bush's paradigm underwent a change and by using Louie Pasteur as an example and where to place him on the previous paradigm, he fell into two categories on both extremes. Hence, the paradigm was shifted to encompass four quadrants that involves the "consideration of use" and "question for fundamental understanding"



Pablo Paredes - 9/28/2010 17:25:46

Summary on Stokes, D.E. – Pasteur’s Quadrant - Transforming the Paradigm

The most relevant outcome of this paper is the introduction of a link between a view of pure basic research and purely applied research via a third category of use-inspired basic research. This view allows a better link between existing and improved understanding and existing and improved technology, showing that there is a need to evolve to a more diverse group of research disciplines, as well as to more complete set of policies.

The historical layout presented in the paper helped understand the evolution of fundamental research, as defined by Bush in the post-war era, passing through bipolar perspectives of research, such as pragmatic and uncommitted, free and mission-related, free and oriented, evolving towards a two-dimensional framework of basic vs. applied, to better fit the Pasteur’s paradigm to finally arrive to a three-prone view of research based on a triplet of fundamental, strategic and directed, finalizing in a proposition of a more complete view of dual trajectories between basic, applied and use-inspired research.

Although the paper does not reach strong conclusions, but reveals the complex and intricate paths of technology development and pure understanding, and clearly shows the importance of use-inspired research as a means to further balance the progress in both fronts, as both technology and understanding need each other to keep growing in a virtuous cycle.


Summary on Kuhn, T – The Structure of Scientific Revolutions

Kuhn’s definition of scientific revolution as the adoption of a new paradigm that replaces an old paradigm helps put into evidence a strong sense of hierarchical community that seems to define the interactions between people devoted to science. Apparently, new, and usually opposing paradigms reflect into huge changes in science. However, it is evident from many examples, such as Einstein, Maxwell, etc., and paradoxical how people devoted to the advancement of understanding are themselves slow adopters of radical changes.

The most interesting perspective for me is summarized in the following quote: “the price of significant advancement is a commitment that runs the risk of being wrong”. It seems to me that the risk factor embedded in proposing new paradigms seems to be too high of a price for many researchers to embrace and therefore the key challenge that only few “pioneer” scientists can embrace, as most of the “professional” scientists are usually immersed in progressing “normal” science, which is less risky and which could even payback a reasonable amount of recognition among peers (considered as a very important intrinsic motivator among researchers).

Although the description made by Kuhn helps understand the nature of scientific revolutions, for me it is more interesting to understand the genesis of such revolutions, i.e. to solve the question how do new paradigms are created? It is for me much more intriguing what is the nature of such “pioneer” researchers. How do they come across new paradigms? Is there a way to “incubate” these pioneers? What are the environmental, cultural, programmatic or not programmatic causes that can bring these new subjects to live?


Aaron Hong - 9/28/2010 17:40:37

In the chapter three, "Transforming the Paradigm," Stokes spends a lot of time rehabilitating our understanding of research by going through the different phases of understanding of research, starting form Bush's "basic research" to finally the Quadrant model. He says that our "understanding" and "use" in research do not have to be divorced concepts and that even technology can be a driver for basic research.

I agree with him. This topic is not that clean cut and that a better understanding will help us make more informed decisions on policy. I'm pretty sure I won't be one of those policy makers, but those who are in academia could one day be on some kind of board or committee. Or may be an advisor of some sort. I particularly agree that those who tried to think one-dimensionally about applications and pure knowledge are trying to reduce irreducible complexity. "It lies rather in the attempt to force into a one-dimensional framework a conceptual problem that is inherently of higher dimension (71)." We need to think more broadly about research.

The essay "The Structure of Scientific Revolutions" by Kuhn talks about the competing paradigms in science. He first starts off with history of science by saying scientific development is "a form of accretion." A cumulative process through which people contribute are compounded. Then he starts discussing the different paradigms but arrives at the conclusion that no paradigm is perfect but: "Which problems is it more significant to have it solve?"


Thomas Schluchter - 9/28/2010 18:09:47

Pasteur's Quadrant

The paper develops a framework for conceptualizing the relationship between scientific research and technological innovation. It does so on the basis of an historic account of the distinction introduced by Vannevar Bush between basic research and its exploitation for practical means, the criticisms of this distinction, and the consequences of the debate for science and technology policy.

I find the framework very valuable. It breaks the unhelpful paradigm of sacrosanct scientific inquiry that needs to shun all consideration of utility to be productive. It stands to reason that it's impossible to predict whether research will lead to technological innovation (or even just recognition), so it shouldn't be the premise for setting up institutional silos. Nevertheless, the timeline shows that in HCI most innovations come out of academic research with an adoption lag of ~20 years. There can be no doubt that the freedom of academic research must include the possibility that the application of this research is beyond the horizon.

The danger of confining all research to areas where practical implications are immediately visible is that innovation will if not come to a halt, at least slow down to baby-steps. Research that is only interested in applicability or even long-term commercialization will not take the kind of intellectual and economic risks typical of the Bohr quadrant. While I agree with the dynamic model of mutual influences between research and technology, it seems obvious to me that the Bohr quadrant drives the quantum leaps in innovation because it is not concerned with any current boundaries. Insofar, Bush's unidirectional model (basic research drives innovation) is too limited. But it has a true core: the understanding in science needs to advance to sustain innovation. Paradigm shifts in understanding can only be achieved when existing conditions don't limit the research.

    • Kuhn: Structure of scientific revolutions

The chapters introduce a theory of knowledge advances in the sciences. The key distinction made for that theory is between 'normal science' and 'paradigm shifting'. Whereas normal science operates under rules and commitments derived from paradigms, paradigm shifts invalidate all these rules and commitments and establish new ones.

The thing that strikes me about Kuhn's analysis is that it seems to apply mostly to single disciplines. This is probably due to the structure that the term 'paradigm' implies. Kuhn mentions biology, physics, mathematics, chemistry and other sciences that have (had historically) a fairly unified trajectory of inquiry. At the critical junctures when old paradigms erode and new ones emerge, the field broadens and becomes less homogenous until a dominant paradigms claims the field. The question is whether an inherently multi-disciplinary field like HCI fits into this model.

The disciplines from which HCI draws are themselves subject to the changes that Kuhn describes. How does a paradigm shift in computer science affect the field of Human-Computer Interaction? It seems rather like interdisciplinary fields might be affected by waves of epistemological changes that spill over from singular disciplines. These waves (think, for example the Structuralism debate of the 60s, or the resurgence of interest in ethnographic methods for cultural description in the 80s) have consequences for how 'science' (in the sense of Wissenschaft) is done, not just how research in any discipline is done.

Because HCI brings in people with different disciplinary backgrounds, it is likely to be, partially, an aggregate of the ways that these individual disciplines conduct their work. The challenge that HCI faces is how to integrate these different ways of conducting work. In that sense, a paradigm for HCI might have yet to emerge.


Siamak Faridani - 9/28/2010 18:27:31

I think the main point of the first article is that science is not simply progressing by an accumulation of knowledge and this linear progression will never end up in interesting results. Rather the history of science goes through a number of paradigm shifts. In these paradigm shifts we see a complete change in the type and goal of the research that is being conducted. He also argues that each paradigm requires a different terminology and one cannot understand or study a paradigm within the framework of another paradigm.

The most productive stage of paradigm shift in science is the Normal Science when the boundaries of science is extended by puzzle solving. Normal science is what seems to be the daily work of a grad student (formulating theories, running experiments, analyzing data and results,...) Achievements in normal science are unprecedented and open ended and they leave room for other scientists to built upon them. Paradigm shifts happen when these scientists start questioning and shattering the boundaries of the Normal Science in their time. Kuhn argues that these paradigm shifts changes the whole scientific community because the new paradigm requires new assumptions that are typically resisted by the science community. This changes the community qualitatively and enriches it quantitatively. Scientific revolutions also matures science.

Each paradigm shapes and defines the boundaries of science and its community. It brings up questions, formulate theories and assumptions and provides definitions and finally it highlights the boundaries of science. Each paradigm starts by random collection of facts, it goes into a pre-paradigm and the a paradigm emerges. A paradigm then grows into different disciplines and professions that finally leads us to new ideas, facts and assumptions that ultimately leads to the paradigm shift.

The author concludes with the nature and necessities of the scientific revolutions. He uses the political revolution analogies and brings up the similarities between the two.

The second article builds upon Vannevar Bush's article and the first article. Although it is in some science a continuation to the first article, unlike Kuhn's article it is more about the strategy for management of science and technology policy. The author starts with this claim that heavy investment in pure science is not efficient anymore.

The author argues that science should both serve as a means to understand phenomena and benefit the society. He builds a framework that splits the science space in four quadrants based on two factors: 1- if that science is pursuing the fundamental understanding of phenomenon 2- If the science can be directly applied to benefit the society. Obviously one quadrant is and empty set. He then provides examples for types of scientists and sciences that appear in each quadrant. Pasteur, Bohr and Edison are presented as examples for each quadrants.

Stokes presents the linear model of scientific development and provides examples of its extensions. Finally he provides the revised dynamic model that consists of three types, the pure basic research, use-inspired basic research, and purely applied research and development. I read and enjoyed the article but I felt that the article was not based on any quantitative or rigorous study and in fact naively formulated.

Luke Segars - 9/28/2010 18:28:54

Pasteur's Quadrant

This paper spends a great deal of time talking about how to classify different types of scientific research. There are apparently a number of views on this issue, ranging from a simple division between “pure” and “basic” research on to more complex multidimensional representations and a spectrum of possibilities between “immediately applicable” and “highly abstract.” Despite their lengthy enumeration of different paradigms, I generally disliked this paper for a number of reasons.

First, the paper never grounded itself in why it was important. It essentially classified itself into the top-right quadrant of the “quadrant model:” the location of topics that are pursued simply for understanding without a particular reason for understanding it. As someone who has performed and is performing research, I felt disconnected and uninterested in the attempts to classify work into one group or another. The paper dedicates a small and inconsequential section at the very end of the paper towards the bond between science and technology with society that does little to improve the situation.

In fact, I question the idea of trying to categorize research into the sorts of classes described in the paper at all. Scientific research has, in many ways, turned from being a way to satisfying curiosity and gain understanding to a deeply bureaucratic method of making money and inflating egos. Perhaps more important than the concepts the paper direct mentioned is some of the assumptions it made: the individual motivations behind research. Why is it important what category a particular research topic falls into? In fact, what is the most important product (or by-product) of research? Is it the outcome of an experiment? There's an excellent quote by Howard Thurman that says, “Don't ask yourself what the world needs -- ask yourself what makes you come alive, and then go do it. Because what the world needs is people who have come alive.” This paper exemplifies the absence this lost life skill in a generally dull and meaningless way.


Drew Fisher - 9/28/2010 18:33:38

The Structure of Scientific Revolutions, Thomas S. Kuhn

This chunk of the book notes that researcher possessing (and thus, assuming shared knowledge of) a commonly accepted base of science is something that until recently did not exist. As a result, by agreeing on many of these "basics," we have been able to push the boundaries of what we know. The book declares that a base set of accurate, assumed knowledge by all researchers involved is really what defines a field as a science.

I attribute this forward progress to getting things so right that there no longer existed current experimental evidence that could disprove a theory. In addition, with more accurate but complicated theories, it becomes harder to push the boundaries of our tests.

Why is this relevant? It shows us that to do research effectively in a scientific field, we must agree on some base assumptions - methods of evaluation, metrics, and so forth. Without these, it is difficult to collect researchers with a common interest or focus. Before we can really do research (generating new knowledge) in a field, we must know what knowledge we've agreed upon and assume. Further, for new research to be considered valid, it must operate within these rules, these assumptions. It is, in fact, it is when these rules change that a science is transformed.


Pasteur's Quadrant - Stokes D.E.

This chapter takes a detailed review of the linear model that was assumed to represent different kinds of research, spanning from basic to applied research. It proposes that Louis Pasteur sits at two points on the scale simultaneously, suggesting that a single-dimension may not be adequate to represent the different kinds of research. The main contribution is a new two-dimensional classification system, with axes of "Quest for fundamental understanding?" and "Considerations of use?" Pasteur now shares one axis with each Bohr and Edison.

I think that this is an interesting approach, but the upshot is largely the same - industrial research (with some exceptions) typically focuses on potential usage (and profit). Academic research tends to focus on fundamental understanding, seeking new models. Industry-funded academic research likely falls inside the overlapping quadrant - Pasteur's quadrant.

Is this a good thing? If adopted by industry and government, it could mean greater funding of Pasteurs quadrant, at the expense


Matthew Can - 9/28/2010 18:45:32

The Structure of Scientific Revolutions (Ch. 1-4, 9)

The goal of this text is to present a conceptual framework for how to analyze the progress of science. In the chapters we were assigned, the author breaks down the development of science into normal science and scientific revolutions. Normal science requires that a field have an established paradigm. It is characterized by the goal of explaining the phenomena and theories provided by the paradigm within the standards and rules of that paradigm. It is not concerned with developing new theories. But, sometimes normal science reveals phenomena inconsistent with the paradigm. And this does give rise to new theories, which can motivate a paradigm shift in the field, thereby creating a scientific revolution. The author argues that such changes are both necessary and sudden.

The author presents several valid reasons for why normal science creates value. For example, the paradigm that forms the foundation of normal science allows researchers to take a set of principles, laws, and theories for granted rather than have to rebuild them from the ground up. This makes the researcher’s job much easier. Additionally, the paradigm narrows the focus of the field and provides the springboard for further research. This guides the efforts of normal science and allows for in-depth examination of specific problems in the field. But, as the author later argues, most of the leaps in the development of science are the products of scientific revolutions. Therefore, it seems that normal science is only valuable to the degree that it is able to beget scientific revolutions. The author does not thoroughly address this issue, but he does state that normal science contributes to the discovery of the scientific anomalies that motivate the development of new theories.

Something I found interesting in this reading was the analogy between normal science and puzzle-solving. The view is that research conducted in normal science is like solving a puzzle because the outcome is anticipated but the process by which it can be achieved is uncertain. The author uses the analogy to provide further characterization of the nature of normal science. What I disagree with is the statement that problems in normal science, like puzzles, have an assured solution. In fact, the lack of a solution or an unexpected solution may reveal a flaw in the paradigm, signaling the need for a new one.


Pasteur’s Quadrant (Ch. 3)

This reading develops a new paradigm between basic science and technological innovation that is more usefully applicable to social, economic, and political needs than is Vannevar Bush’s model, in which large investments in pure, basic science will generate the technological innovation necessary to ensure a successful economy. Rather than classifying research on a one-dimensional basic-applied spectrum, the new paradigm classifies research in two dimensions: (1) Does it give consideration to use? (2) Is the purpose fundamental understanding? This taxonomy gives rise to a dynamic model of the relationship between basic research and technological innovation. In that model, research and technology develop incrementally in parallel, loosely coupled through achievements in use-inspired basic research.

This text is particularly relevant to HCI because in this field more than in others, the issue of how applied research should be versus how theoretical or fundamental it should be plays a role in the sort of research that is undertaken. Because of its nature, HCI lends itself to applied research. But does that necessarily mean HCI research should be primarily applied?

The author notes that concerns over economic competitiveness have led some to focus on the commercial application of new technology. The text quotes Erich Bloch and David Cheney as having said, “Technology that remains in the lab provides almost no economic benefits.” As the author explains, while the US may lead the world in basic science and technological innovation, it is less successful at “converting new technology into products and services that meet the test of the market.” This raises some important questions. Assuming commercial application is a desirable thing, whose responsibility is it to achieve it? Certainly it seems nonsensical (and perhaps even conflicting with the goals of science) to expect researches to conduct their research with a product or market in mind. Somewhat more reasonable is that funding agencies allocate funds based on an agenda influenced by market considerations. But probably the only efficient way to commercialize technology is through industry-backed research, where the goals are compatible. Another way to look at this issue is to consider the distinction the author described between technology innovation and technology exploitation. Separating the technology from its application gives the researcher more flexibility in his research by removing applicability constraints. But, as the author notes, this separation means that market forces do not directly inform the trajectory of technology. Perhaps the observation that technologies developed in the US are exploited first in other countries can be explained by this drawback to distinguishing technological innovation from its application.


David Wong - 9/28/2010 18:46:47

1) The "Structure of Scientific Revolutions" discusses the nature of scientific revolutions and how they do not coincide with the general public's view of scientific development. The paper argues that scientific development, rather than being cumulative and constructive as the general public may assume, is a chaotic process where assumptions from one paradigm are replaced by another paradigm. The "Transforming The Paradigm" paper discussed how the definitions of research have changed since Bush's "As We May Think" essay. The paper describes in detail the nuances in defining basic and applied research.

2) I thought that the "Structure of Scientific Revolutions" paper, although not directly related to HCI, gives an interesting perspective on research as a whole. Rather than focusing on problems that are "guaranteed a solution", researcher can also challenge the very paradigms that give structure to their research. The paper does not directly espouse engaging normative research or trying to challenge existing paradigms, but gives the positive and negative effects of each. Rather, the paper helps open a researcher's eyes to the higher-level perspective of how one's research is impacting both the specific field and the general public. In that sense, I think the paper is very useful.

I thought the "Transforming The Paradigm" paper had no real added value, except perhaps to allow a researcher to better petition for a grant from an institution. It seemed like an attempt to categorize research for the sole purpose of statistics and reporting by an organization. The paper even stated that "everyone knows the linear model of innovation is dead" and that "it still lives on in parts of the science and policy communities". Essentially, it seemed more like an exercise in categorizing research, when the researchers themselves can already report what they are doing.

3) The "Structure of Scientific Revolutions" paper, for the most part, had a very sound argument supported by an extensive, and impressive, recount of historical events. The only aspect of the paper that I felt should have been emphasized more is the fact that normal research in many ways leads to paradigm shifts. Researchers engage in normal research in order to make incremental steps towards better understanding their paradigm. Only with the better understanding can one begin to question the paradigm.

The "Transforming The Paradigm" addresed a problem that, to me, was not highly motivated. While it is true that a well defined framework for defining research can help organizations better allocate money, it does not seem like a pressing issue for research as many institutions can create definitions unique to their needs. Moreover, the framework that they suggested, in my opinion, did not offer any real insight into the topic. Although they predicted a new type of research, the lower-left quadrant, the value of identifying that quadrant, in my opinion, is not very high. I am, however, taking an applied research perspective in evaluating their work.


Aditi Shrikumar - 9/28/2010 18:53:51

For Thomas Kuhn, the central problem seems to be that if we take science to be some incremental accumulative process of gaining knowledge about the natural world, how can science about an existing phenomenon change? How can scientific "revolutions" occur? His answer seems to be that science is not a collection of truths, but a widely-accepted "paradigm" that fits the current observations (i.e. a theory) and that if you accept this view, then scientific revolutions are simply a change of paradigm. This makes sense, but I don't think it required as many pages as he took to say it.

His verbose essay seems to suffer from an outsider's unfamiliarity with science, which is odd because he had a PhD in Physics. The "new" view of science as a paradigm or theory about how the world works that Kuhn describes is entirely acceptable - but it's presented in competition with a straw man. This view of science is new only to science historians, and not to scientists, who would never elevate scientific theories to the status of fact. Richard Feynman, a contemporary of his, and a Nobel-prize-winning physicist, had no such conceptions and wrote many books about science for the layman in which he demonstrated this. Nevertheless, I never managed to satisfy myself that I understood Kuhn entirely, so I may be wrong.

D.E. Stokes, in Ch3 of "Pasteur's Quadrant", discusses the different ways that policymakers have classified reasons for scientific research. Stokes explicitly tackles one subjective aspect of science that Kuhn neglects: that the very directions in which scientific curiosity applies itself are influenced by political, social and historical forces, most importantly, the availability of funding.

He introduces increasingly refined ways of categorizing scientific research based on initial motivations and end results. The name "Pasteur's Quadrant" results from one of the categorization scheme based on initial motivations. In this scheme, the desire to find applications is on the horizontal axis, and the desire to gain knowledge is on the vertical. Pasteur occupies the top right, with both desire for application and for knowledge acquition. Bohr occupies the top left, with no thought for applications, and Edison occupies the bottom right, with no motivating desire for knowledge, but a drive to find applications.

His main point, which i agree with, and which has proven accurate in retrospect, is that policymakers must recognize the importance of "use-inspired" basic research - and no longer distinguish sharply between applied and "pure", funding one and not the other. He stresses that innovation and useful applicable results will come about by allocating funds based on the potential of research to produce results that could potentially be used to solve society's problems.