- 1 Bjoern's Slides
- 2 Extra Materials
- 3 Discussant's Materials
- 4 Reading Responses
- 5 Shiry Ginosar - 10/2/2011 23:54:43
- 6 Valkyrie Savage - 10/3/2011 19:23:38
- 7 Hanzhong Ye (Ayden) - 10/4/2011 13:27:12
- 8 Steve Rubin - 10/4/2011 16:07:32
- 9 Laura Devendorf - 10/4/2011 17:35:01
- 10 Yin-Chia Yeh - 10/4/2011 20:22:04
- 11 Galen Panger - 10/4/2011 21:05:21
- 12 Donghyuk Jung - 10/4/2011 21:58:36
- 13 Suryaveer Singh Lodha - 10/4/2011 22:26:22
- 14 Yun Jin - 10/4/2011 22:30:53
- 15 Hong Wu - 10/4/2011 23:15:20
- 16 Peggy Chi - 10/5/2011 1:27:41
- 17 Ali Sinan Koksal - 10/5/2011 3:50:42
- 18 Derrick Coetzee - 10/5/2011 4:56:29
- 19 Vinson Chuong - 10/5/2011 8:24:09
- 20 Manas Mittal - 10/5/2011 8:37:16
- 21 Allie - 10/5/2011 8:37:45
- 22 Alex Chung - 10/5/2011 8:42:26
- 23 Apoorva Sachdev - 10/5/2011 8:50:54
- 24 Jason Toy - 10/5/2011 8:58:47
Shiry Ginosar - 10/2/2011 23:54:43
In this section of his book, "The Structure of Scientific Revolutions", Kuhn discusses the role of scientific paradigms in directing normal research, and the nature of the revolutions from which these paradigms emerge. Paradigms define the questions, the methods and the possible solutions that are accepted as being scientific in a specific domain. According to Kuhn, they are necessary for the accumulative course of normal research which strives to fill in the gaps in the paradigm's coverage using problem solving methods and motivations which Kuhn likens to puzzle solving. Contrary to the course of normal research, transition from one paradigm to another often happens when the first cannot explain anomalities (?) in nature that the latter can. As in this case two paradigms often present a different view of the world, they cannot logically coexist and the newer must replace the former via a revolution.
This view of research as a bimodal activity composed of large leaps of paradigm shifts, or revolutions, and small accumulative steps of normal research designed to fit our view of nature to the box defined by the paradigm is an insightful and instructive one. As an incoming PhD student, much of one's effort is spent considering different areas of research and attempting to grasp the set of questions and methods unique to each. As an outside observer of a body of research, it is often difficult to understand how researchers in that field make choices about the questions they solve and the ways they go about it. Many times it seems like the basic assumptions can only be explained by what is straightforward or reasonable to do based on the path that the particular field has taken thus far. It is interesting therefore to take these assumptions with a grain of salt, and keep in mind that they are indeed a result of the currently accepted paradigm in the field. Who knows? Perhaps by the time we are done with our thesis someone will come up with a new paradigm and a revolution in the questions and methods will take place. Perhaps this someone will even be one of us.
This time, I think I will refrain from voicing a criticism of the article. Instead I think I will try and use Kuhn's vision as a lens through which I can inspect the research around me in hopes it can help me in setting my own course.
Valkyrie Savage - 10/3/2011 19:23:38
Science does not happen in a vacuum. Usually. Sometimes the best science comes from nowhere: great leaps that don’t so much build on what is known as redefine it.
This paper covered a lot of ground, but there were notable things that I will take away from it (aside from the obvious comment that the author is quite long-winded). First, I liked his comparison of ‘normal’ science to puzzles: it seemed well-suited as a metaphor. It’s been an interesting transition to the place that we currently are, where science is, in fact, defined by rules. It makes sense, of course, that this is the case: in such an age where there are innumerable scientists it would be difficult to understand advances made without context which bent rules. It reminds me somewhat of Principia Mathematica, which was a great work that no one was well-able to decipher until its own structure managed to disperse sufficiently. Scientific revolutions can’t be understood until they are understood. I am a strange loop. (Hat tip to Doug Hofstadter.)
I wasn’t convinced by the large number of enumerations made by the paper. As it was largely what I would call a thought experiment, it didn’t seem necessary to slice and dice everything as “tidily” as was done: three types of phenomena, three types of normal science, four parts in the network of a scientist’s commitments. He’s adding rules and counters and measures to describe things that are, by the very definition he gives, rather averse to those sorts of structures!
Hanzhong Ye (Ayden) - 10/4/2011 13:27:12
Reading Response for Oct 5th:
The Structure of Scientific Revolutions.
The reading for Wednesday class is a long discussion talking about the revolution of scientific research in a historical aspect. The book starts with an introduction of a role for the history in the overall process of human science development, and then discusses specifically about people’s route to normal science. Following with a discussion on the intrinsic natural of normal science, which to some extend is very alike as a procedure of puzzle-solving. I strongly feel that this feeling of puzzle-solving is actually the most exciting part and also the strongest motivation for human to explore in science, because of the born curiosity of human beings. The author also talks a lot about the process of scientific revolutions in terms of their nature and necessity. From my point of view, I believe the revolution itself is an essential part of scientific research, without which the process of scientific research would have been significantly slowed down. Just as in political field, a revolution not only brings fresh thoughts and methodology, but also brings opportunities of reconstruction and giant leap with improvements.
-By Ayden (Oct 4th, 2011)
Steve Rubin - 10/4/2011 16:07:32
In "The Structure of Scientific Revolutions," Thomas Kuhn centers his discussion on the notion of paradigms and paradigm shifts (a phrase that he invented in this book, Wikipedia tells me). The basic principle is that some set of theories makes up the current paradigm in a field, and normal science is done with respect to that paradigm: to reveal the nature of things in that paradigm, to compare fact with predictions from the paradigm, and to better articulate the details of the paradigm. Scientific revolutions happen when a "better" ("more correct," if you will) paradigm comes along to replace the old one. Such shifts, Kuhn argues, are not incremental changes, but major theoretical shifts.
The essay notes that some questions, like "how do we cure cancer?" and "how do we achieve world peace?," do not fit into the mold of normal scientific inquiry. This is my favorite point that he makes. These questions are not well-constrained, and because we cannot be sure whether these questions are answerable, we can not create a mental framework of what an answer to these questions would look like. While these questions are still noble goals, progress toward answering them would take the form of smaller "puzzle-solving" tasks. By considering problems as puzzles to be solved, we are assuming that there are possible solutions, and that the things we would consider solutions are constrained. This framework makes the job of the "normal" scientist much more focused.
The actual examples provided by Kuhn--oxygen, heliocentrism, relativity, etc.--are useful initially, but eventually bog down his argument. By the end of the reading, we all knew about the dramatic paradigm shifts involved in these theories. In fact, even with my American public middle school and high school education, I knew about these revolutions a priori to reading this paper. Maybe at the time that it was written, these revolutions were less widely known and therefore more illustrative, but I doubt that many people have read this essay that were not already well-versed in the big events of science history. The essay could have been half the length if Kuhn did not constantly appeal to science history--once was enough.
Laura Devendorf - 10/4/2011 17:35:01
The reading discusses the history of science and the paradigm shifts that change the conception of what is acceptable research.
Having budgeting time for the usual HCI readings, I was unable to devote the full depth of reading which this article is entitled. I think the subject is important and interesting and I plan on reading it again in more depth. I particularly enjoyed the section on normal science as puzzle solving as it got me thinking about the roles of science and art and rekindled a discussion we had in class about design and art. In the puzzle section the author mentions how science looks at problems with expected outcomes and how the conclusion is getting that outcome in an a new way. Where design holds more closely to this idea, I feel that some forms of art attempt to shed light on problems that may not have solutions. Often, I find that artists are more willing to tackle broad problems for which no feasible solution exists. I also enjoyed the comment that "the price of significant scientific advance is a commitment that runs the risk of being wrong."
Yin-Chia Yeh - 10/4/2011 20:22:04
Today we read part of the book “The structure of scientific revolution,” written by Thomas Kuhn. The gist of the reading is that as we might believe that the progress of science is a continuous process of adding pieces of knowledge into known knowledge base, it is not the case. Kuhn proposed that the progress of science is separated to periods leading by scientific revolutions, or paradigms. They way researchers think and work is directed by paradigms of their time. Paradigms set the game rule and researchers were the “puzzle solvers.” Researches working based on a paradigm is called normal science. Three types of works in normal science are as following. First, determine some significant facts, ex, measuring some universal constants. Second, match facts with the prediction of paradigm. Third, refine the paradigm to make it more precise. The advantage of normal science is it guides researchers to think deeply at certain problems best described by paradigms. The disadvantage of normal science is that it causes researchers overlook certain problems cannot be explained well by a paradigm. When more and more conflicts of a paradigm occurred, or when the problem set defined by a paradigm is exhausted, researchers start to look at things they neglect before until a new paradigm replaced the former one. Successive paradigms are by definition different and the difference is irreconcilable, otherwise it won’t cause a major change of how researchers think and work. I think this reading suggests two very different research types, either working on current paradigm to extend human knowledge base or developing something new to create whole new path of research. The reading provides a lot of samples on basic science. I would like to learn more about how it works in computer science. One question I have in mind is that does the three different types of normal science work applied in computer science? To me they seems not fitting well to computer science. I think it’s probably because computer science is applied science which the motivation of research is not the same as that in basic science. On the other hand, paradigm shift seems applied in computer science, if the case that some publication changed the direction of entire community counts. There are some examples in my mind like SVM in the field of classification, particle filtering in the field of object tracking, and Markov random field in the field of natural language processing. But I am not 100% sure if such breakthrough can be marked as revolution because the examples in the reading are so huge discovery.
Galen Panger - 10/4/2011 21:05:21
Wednesday’s reading comes at an interesting time for me. In my political behavior seminar, the fourteen or so of us Ph.D.s (mostly political science; I am the sole I Schooler) and professors sits around a table discussing seminal and current pieces in the realm of explaining the psychology of mass politics. It’s a realm filled with amazing contributions—from Milgram’s classic study of obedience to Edelman’s elucidation of symbolism.
But it’s a strange seminar because the dominant paradigm in political science continues to this day to be the rationalist (neoclassical economic) paradigm, which argues that man can be modeled essentially as a self-interested calculator. Other elements of human experience, such as emotion, are left out of the model mostly because they don’t fit.
The rationalist model says that to be human is to have a ranked order of transitive preferences (meaning they stay in the same order and don’t shift around based on the context), and the way we know people’s preferences is through the (assumed to be consistent) decisions they make. For decades, critics of the rationalist model have poked holes in this idea of human decision making, leading to “bounded rationality” models and, more generally, cognitive science which generally takes an analytical, rather than emotional, view of human nature.
Neuroscientists can confirm that which psychologists generally and social psychologists, perhaps, in particular, have always known—that we are first feeling beings and second thinking beings.
But try telling that to political scientists, many of whom continue to believe that the rationalist model is, while incomplete, most of the story of human nature. And from that frame of reference all questions deemed interesting problems, as well as all possible answers, flow. So, rather than a “focus on the user” sort of approach that HCI might take to political and mass voter behavior, political scientists take a top-down, judgmental approach that bemoans the “irrationality” of voter behavior and says voters are simply too ill-informed. So the mystery of why democracy “works” remains, as does how it works. HCI researchers by now would probably be on to the question of aggressively designing new ways to make it work better.
There would obviously be dark rooms filled with depth cameras and projectors shining onto every surface.....
Donghyuk Jung - 10/4/2011 21:58:36
The Structure of Scientific Revolutions
The Structure of Scientific Revolutions (SSR) consists of 13 chapters. As the introduction, 1st chapter represents historical means of telling scientific history in order to introduce new possibilities for theory of scientific knowledge.
2nd chapter describes how paradigms are created and what they contribute to scientific (disciplined) inquiry. The author also mentioned ‘Normal Science’ and he describes it: “Research firmly based upon one or more past scientific achievements, achievements that some particular scientific community acknowledges for a time as supplying the foundation for its further practice.” In addition, he argues "The successive transition from one paradigm to another via revolution is the usual developmental pattern of mature science."
In chapter 3 to 5, he describes roles and characteristics of normal science as puzzle solving as well as uniqueness and functions of paradigms, which are exemplified by major scientific incidents. According to the author, doing research is essentially like solving a puzzle because it generally has predetermined solutions and rules.
Lastly, he explains why a paradigm change should be called a revolution and what the functions of scientific revolutions in the development of science are from chapter 9 to the end. He believes that a scientific revolution that results in paradigm change is analogous to a political revolution. In this sense, a scientific revolution is a noncumulative developmental episode in which an older paradigm is replaced in whole or in part by an incompatible new one.
From my understanding, the main idea of his philosophy is that changes and improvements of scientific knowledge are revolutionary. So, he denied that scientific knowledge evolves cumulatively from verifying observations and experiments. Overall process of paradigm shifts is following: Pre-paradigm -> new paradigm -> normal science -> challenges -> emergence of competing paradigm -> scientific revolution -> new normal science -> new challenges
Suryaveer Singh Lodha - 10/4/2011 22:26:22
Kuhn discusses the value and creation of paradigms in the practice of science. He develops his argument through a series of major examples of paradigm shifts in the history of science, such as the replacement of Newtonian mechanics with Einstein’s mechanics. He also analyzes the progress of science. He breaks down the development of science into normal science and scientific revolutions into finer paradigms. 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 is inconsistent with the paradigm which gives 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 essential and sudden. The analogy between normal science and puzzle-solving is interesting. 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 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 job of researcher much easier. Additionally, the paradigm narrows the focus of the field and provides the basis 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. To relate to theories, their range of application must be restricted to the exceptions which they cannot explain and to precision of observation with which the experimental evidence in hand already deals. Without commitment to a paradigm there could be no normal science. Furthermore, that commitment must extend to areas and to degrees of precision for which there is no full precedent. If it did not, the paradigm could provide no puzzles that had not already been solved. It is important to note that the price of significant scientific advance is a commitment that runs the risk of being wrong
Yun Jin - 10/4/2011 22:30:53
Reading response of the structure of scientific revolutions：This paper is an analysis of the history of science. It triggered an ongoing worldwide assessment and reaction in — and beyond — those scholarly communities. In this work, Kuhn challenged the then prevailing view of progress in "normal science." Scientific progress had been seen primarily as a continuous increase in a set of accepted facts and theories. Kuhn argued for an episodic model in which periods of such conceptual continuity in normal science were interrupted by periods of revolutionary science. During revolutions in science the discovery of anomalies leads to a whole new paradigm that changes the rules of the game and the "map" directing new research, asks new questions of old data, and moves beyond the puzzle-solving of normal science. And I have some problems with the paper, I defend or criticize the following inter-related views propounded by thomas kuhn: a) Paradigm is the basis of a well-developed or mature science b) Normal science is mainly puzzle solving c) Dogmatism is useful in scientific research d) Science educations, thru textbooks, is initiation into a paradigm e) Paradigm changes as scientific revolutions f) Since competing paradigms are incommensurable, paradigm choices are never categorically or completely objective.
Hong Wu - 10/4/2011 23:15:20
Main Idea: The paper introduced the nature and history of normal science and then described the revolution of scientific science Interpretation: The paper first reviewed the history of normal science and then invested the nature of normal science. In the normal science, the author summarized what normal science is consisted. First is class of facts of nature of the things. Second is the fact to predict from theory. When we try to prove a theory by experiment, it is the second kind. Third consists experiments and oversevation exhausts. The paper further describes the nature and necessity of scientific revolutions. This is based on the author’s understanding that no paradigm will solve all the problem. At the same time, any two paradigms can solve at least problem. To me, it is very important to understand the difference and relationship between nature science and scientific science. It will help us to improve the method we use in our future career.
Peggy Chi - 10/5/2011 1:27:41
In these book chapters, Kuhn analyzed the history of science and proposed one of the important and classic discussions, "paradigm shift". The science under one paradigm is called "normal science", and the efforts within are like puzzle-solving. When revolution happens, the paradigm shifts and brings people new ways of thinking.
The concept of "paradigm" is so intriguing and profound that pushed me to rethink the HCI research field. When radical ideas such as human computation, ubicomp, cloud computing, crowdsourcing, tangible interfaces, or software agents appeared and challenged people's mindset, are we accumulating our knowledge to shift the overal paradigm? Should we accept what we've known and agreed based on the existing research, such as the studies of Fitts' Law, machine I/O, direct manipulation, etc.? Should we put on efforts based on such basis? What do we commonly believe? HCI is such a complicated, interdisciplinary area that cannot be defined and described easily in few sentences. One interesting fact is that, currently it's even hard to say what the "bible" or common textbook of the field is, which might imply that we are "in" the new revolution, or in the "pre-paradigm" stage, moving toward a new age.
Ali Sinan Koksal - 10/5/2011 3:50:42
Kuhn's "the Structure of Scientific Revolutions" draws the line between "normal science", which consists essentially in "puzzle-solving", that is, attacking problems for which a solution is expected to be found using the techniques that are suggested by the well-established paradigms in a field, at a given period of time. Kuhn states that, while paradigms help structure research and allow it to explore some of its aspects very deeply, shaping one's mind based on rules deduced from paradigms will not result in ground-breaking progress. For true progress to be made, a "paradigm shift" should occur: a new paradigm, typically in conflict with a former one, should emerge and be widely accepted, in ways that might not involve only logic, but dispute and debate, as in political revolutions.
This work was very interesting to read (I've been meaning to read it for some time already) and definitely offers a new point of view for describing the nature of science and how it evolves. Science is not achieved by continually accumulating theories and knowledge, but new paradigms replace older ones as they offer better explanations for phenomena seen as anomalies in older frameworks. I agree that strictly abiding by given paradigms may hinder our capacity to devote enough attention to these anomalies that might be the sign of the fallacies of the prevailing paradigms. This reminds of our reading on design models, during which the restrictive nature of models had also been considered.
This piece of work has surely created a great amount of criticism and discussions. What Kuhn considers as very idealistic and improbable, the continuous progress of science, is defended by Karl Popper in his work. Popper values "falsifiability" and according to him, science makes progress by eliminating points of views that are falsified. Another notable reaction to Kuhn's work is its integration of non-scientific factors into scientific progress, which is also understandable.
Overall, this paper is definitely very influential, and leads one to think about which questions researchers should investigate, and how that should be done.
Derrick Coetzee - 10/5/2011 4:56:29
This 1962 work by Thomas Kuhn described a novel conception of how science progresses over time, in which a field is initially organized into competing camps which conceptualize the same experimental results differently, but then are eventually unified into a common paradigm, which in turn is subsequently replaced by other paradigms in a disruptive scientific revolution. Between these revolutions, inquiry is dominated by "normal science" which explicates the object of study in detail in terms of the dominant paradigm.
His theory of scientific development is illustrated by numerous historical examples. In a number of cases he emphasizes that historical adherents to invalid theories still did valuable scientific work and held views compatible with available evidence - and conversely, that they held views fundamentally at odds with future paradigms, necessitating a revolution rather than an incremental accumulation of knowledge.
Although a persuasive account of how science progresses, the essay has a number of limitations. First, it supposes that once a paradigm has been established that competing theories vanish, when in reality they likely continue to be held by small groups who may still function as an independent research community. The interplay of dominant and minority paradigms is not explored. Another limitation lies in the supposition that the historical progress of science is a good predictor for future evolution, but it seems difficult to ascertain that science is not moving as a whole towards having less (or more) revolutions, as opposed to incremental discoveries.
Finally, like many works of its era, Kuhn's work is dominated by references to the development of Western science, failing to draw historical examples from other cultures such as known early work by Chinese mathematicians. This is the most concerning omission because the model is primarily a sociological one, and so vulnerable to differences in culture. For example, perhaps revolutions would be more gradual and less dramatic in a culture that values group harmony highly (as in many collectivist societies).
Vinson Chuong - 10/5/2011 8:24:09
Kuhn's __The Structure of Scientific Revolutions__ offers a model of scientific progress, of how competing ideas coalesce and converge into a sharp focus, of how that focus is refined and anomalies are uncovered, and of how new ideas arise to repeat the cycle.
Kuhn motivates his model by calling out the common misconception that scientific progress is linear---new discoveries and insights merely add to the existing body of ideas---and thus the "science" of today is a refined and fleshed-out version of the "science" of the past. No, Kuhn asserts that scientific progress is much more like a tree in that multiple branches are deeply explored and most are pruned, leaving a single path, or paradigm, that is held to be true. As inconsistencies accumulate and new ideas arise to address them, the current paradigm may be displaced or replaced by an entirely new one, a "scientific revolution", and current theories are either adapted or discarded. Kuhn's model offers a not only a way to describe and analyze the process of scientific progress but also to objectively compare past and present paradigms.
However, in presenting this model, Kuhn makes many controversial assumtions. He assumes that once a science has matured taken on a paradigm, inconsistencies and anomalies will be rejected or ignored; change will only occur after the build up of these anomalies reaches some critical mass, and an alternative paradigm is presented. He also assumes that if and when a new paradigm is accepted, it will be "incommensurable" with the old---that is, it will completely redefine the science. Kuhn seems to exclude the possibility of small and continuous revisions to a paradigm---that is, in his model, scientific progress is like a tree with discrete branches as opposed to a continuous space. I believe more discussion is needed to justify these assumptions.
Moreover, there seems to be a gap in Kuhn's model between when a body of ideas is accepted and when a paradigm is formed (that is, before a science matures). How does progress look in sciences that are not yet "mature"? Kuhn draws most of his evidence from well-established areas like physics and chemistry; are these reliable representations of what science is, or is Kuhn being too specific?
Overall, I believe that while the ideas brought up in this book are thought-provoking, the model presented is far from complete.
Manas Mittal - 10/5/2011 8:37:16
Reading tries to device a taxonomy for scientific work, and, based on a historical perspective, attempts to describe the patterns of how research is done, and how theories come into vogue and go out of vogue.
I wonder where HCI stands as a field in terms of Kuhn's taxonomy. Are people who find out the constants (perhaps like the perceptual response time to stimuli), those who find out fundamental constants (like 'G' in case of Newton, can't think of an HCI parallel), and those who develop theories (fitts law?) working on each others work?. I think one way to measure this 'building' on each others work would be citation index, and perhaps this is why researchers and program committees like that people cite work - it makes you part of the 'community', and accept the common values of the community.
One example where this system didn't work was the AI community. The popular beliefs (and influential leaders) accidently severely impeded the field when they mistakenly claimed that a particular technique won't/can't work (specifically, Minsky & Papert 'proved' that neural networks can't XOR in the 60's). The community accepted this, and funding for this research was cut. What was needed was, perhaps, a scientific revolution, but it took many years for that to happen. In the 80's, the complete field was shaken and has become statistical (neural network, and bayes net) in style rather than based on qualitative theories of how to think, i.e., machine learning has become synonymous with Artificial Intelligence.
The book also talks about how scientific revolutions happen, and how such revolutions happen based not just on rational, logical arguments, but also on a leap of faith.
Allie - 10/5/2011 8:37:45
Kuhn's "Structure of Scientific Revolutions" gives a survey of the development of scientific methods and paradigms in recent Western history. He asserts that traditional education is rigid, and scientific breakthroughs must overcome the frigid conceptions of education. The individual scientist himself needs to remind himself not to take paradigm for granted, because then he would no longer build anew by starting from first principles to justify each concept. Kuhn covers 3 normal foci for factual scientific investigation, which govern similar characteristics in the evolution of the sciences: 1) the revealing of the nature of things in orer to determine significant fact; 2) matching of facts with theory, in particular facts that can be compared with predictions from paradigm theory; 3) articulation of the theory themselves, or fact-gathering. Since nature is too complex to be explored at random, a methodical exploration is important to the continual development of science.
Kuhn goes on to say that although certain outcomes can be expected in scientific research, the way to achieve that outcome is like a problem that must have alternative solutions. The scientist himself must extend the scope of the world by scrutinizing details in nature. He cites quantum physics (Einstein) and mechanics (Knewton) as conflicting theories. Generally, an out-of-date theory can be viewed as a special case of its up-to-date predecessor, with credence given to the existing theory in the process of transformation.
Some theories require consent of the community in order to gain foothold: Knewton and Maxwell's theories are exemplary of this. Since no paradigm ever solves all the problems it defines, in choosing between paradigms, the question comes down to the most significant problem to have solved.
Kuhn's paper relates to our discussions in HCI this semester thus far. While the evolution and discoveries made in science is man's statement of the world as he sees it; human-computer-interaction research focuses more on his reactions and role as an actor in partnership with technology. The methods of HCI research are akin to traditional scientific paradigms, although less so in the sense that scientific discoveries tend to be definitive until a better theory replaces its predecessor, whereas preferred HCI methods are more contingent on the variables surrounding the particular type of research administered.
Alex Chung - 10/5/2011 8:42:26
Summary: Science is informed by a series of paradigms. Most of their work is focused on validating the current principles in their respective fields (see physicists). Scientific revolution comes when the overwhelming facts prevent the naysayers from turning their heads.
Is HCI research like normal science?
While normal scientific research is similar to puzzle solving where the outcome is known ahead of the experiment, HCI research often deals with an unknown or only partially known situation. It certainly does not attempt to force nature into the “preformed and relatively inflexible box” that the scientific discipline supplies. On the contrary, HCI research reaches across multiple disciplines both in concept and in practice.
However, from earlier readings, there exist design theories and design philosophy works that have gained traction among interaction designers. Although some of their works seem meaningless at the time, theorists like Vannevar Bush and researchers such as the guy behind Skinput produced a diverse intellectual foundation of knowledge and insights that future interaction designers could build on. In this regard, this is similar to Kuhn’s description of a paradigm that a body of “intertwined theoretical and methodological belief that permits selection, evaluation, and criticism.” After all, there are already many skilled practitioners of interaction design such as Wilson from Microsoft demonstrating the interactive virtual room to Wellner creating the digital desk using camera and paper.
HCI is becoming a mainstream discipline to understand the philosophy of design practices and to develop design methods and techniques for practicing designers. It has certainly gained its status from solving the problems of human interacting with the ubiquitous computer devices. Yet there is still plenty of “mopping-up” to do. For example, we are still struggling to quantify the results in order to compare two interactive systems. Although we’ve been introduced to Fitts’ law and other measurement principles when studying input models such as bubble cursor, many subsequent papers skipped the data gathering and analysis.
Finally, I don’t believe HCI research is like normal science. Traditional science is about exploring the natural phenomena, discovering the guiding principles that explain intricacy. On the other hand, HCI research involves the discovery of how we interact with man made objects in order to improve the next generation of gadgets or devices. Current HCI research focuses on engineering improvement rather than exploring the fundamental behind natural tendency toward one design more than the other (similar to the study of directness). Unlike the traditional science, HCI research does not limit oneself to a boundary. The new discipline could be searching for its soul or charting a new multi-disciplinary study. Either way, HCI has the potential to be a revolutionary science requiring a new pair of lens to value its worth and contribution.
Apoorva Sachdev - 10/5/2011 8:50:54
This week’s readings were parts of book by Thomas S. Kuhn. He talks about the structure of Scientific Revolutions. He attempts to answer a few of the conceptual questions that revolve around scientific revolution and how it should be documented historically. He goes on to describe how the field of research and the kind of research that is being done has changed over the years.
The author in the beginning presents several arguments about how scientific history should be documented and whether science is developed “by the accumulation of individual discoveries and inventions” or as society as a whole instigated by changing intellectually circumstances and possibilities. It is hard to pinpoint exactly who made certain discoveries and when, so it might be more relevant to look at school of thoughts and ideas rather than being lost in determining the specifics. Also, it is harder to distinguish science from myths and beliefs as over the years our impressions of the scientific world have evolved. I agree with the viewpoint that instead of studying contributions of scientist like Galileo and Newton in relation the modern science, their contributions should be studied in the context of their contemporaries.
However, some points that the author presents are rather interesting but make not necessarily be true. For instance, he claims that in recent years scientist have spent more time trying to prove a known outcome rather than necessarily finding new outcomes. (“The end result is not as interesting as the method used to achieve it”, thus science has become somewhat like “puzzle-solving”. Although, this is true in some instances, I feel stating that this puzzle–solving approach has limited research or restricted scientist in proving only what is already known is not entirely correct. The author also seems to stress on the fact that, in this age we are solving problems of little novel, conceptual or phenomenal significance. This is debatable as very recently scientist have claimed to have found neutrinos that are supposedly faster than light and this is a huge discovery that would change a lot of the foundations of our current understanding of particles and physics. (http://www.nytimes.com/2011/09/23/science/23speed.html)
Overall, I agree with some of the points the author has presented in that scientific revolutions could be looked as “non-cumulative developmental episodes in which an older paradigm is replaced in whole or in part by an incompatible new one” which forces scientist to reconsider their perspectives and ultimately leads to new discoveries and hopefully a truer understanding of the world around us.
Jason Toy - 10/5/2011 8:58:47
The Structure of Scientific Revolutions
"The Structure of Scientific Revolutions" is a different viewpoint on what constitutes science and scientific revolutions.
This paper is a new intellectual framework that says that rather than being built upon laws that govern the natural world, science is built upon paradigms in each field which answers questions, not fully, but the best. Paradigms are the way we judge methods and results to be scientific. They are only removed when they are completely discounted, but until that happens, all science is built around proving and refining the ideas that they propose. Since there is no absolute, we should judge scientists' ideas based on what their peers were thinking at the time rather than what we believe is true today. In addition this paper draws analogies to political revolutions and approaches similar to those detailed in "Dilemmas in a General Theory of Planning". This article points out that competing paradigms are like a choice between incompatible modes of community life. Because of this, the evaluative procedures of normal science cannot be applied. As a result, scientists have constrained themselves to problems they can solve, rather than the evil problems that Ritel and Webber discuss. This paper may influence what experimenters consider to do future experiments about by having them think about or ignore paradigm constrictions.
One thing the paper does good is to acknowledge the fact that paradigm shifts do not have to prove old paradigms wrong. In the case of Einstein and Newton, Einstein's paradigm was just a generalization of Newton's. In addition it is important to acknowledge that scientific paradigms do not explain everything perfectly today. What we have written in textbooks is our best guess at how the world works, rather than being "natural" laws of the world. However the paper fails to define science. I argue that inventions should be considered science. If this is true, then I disagree with the fact that one person cannot make a difference, providing the counter example of Thomas Edison. The story of Edison and Tesla's battle over DC and AC as an analogy to a paradigm being pruned off. Tesla lost that battle and was shunned as a mad scientist. However, today, we use AC in our homes, proving that a paradigm shift is not as black and white as described in papers.
I did not like the style of this paper. It seemed to adopt the strategy of asking a question, proposing its answer for most cases, and then changing its mind in the last paragraph of a section. For example, the paper stresses the importance of rules of a paradigm in everything from "scientific" methods to accepted results. However in the end, it accepts that paradigms can be used in lieu of rules if necessary. Another example was the idea that revolutions involve displacing old paradigms. For a large part of the paper, it appears that the author is of the opinion that paradigms must displace previous ones. However, he finally comes up with the example of Einstein and Newton to counter this idea. Another thing I disliked was that the author considered only examples from hundreds of years ago. This leads to the question, has there been no real life-changing science being done in the last couple hundred years? The world we live in today seems to prove otherwise. How does something like computer science fit into the realm of science? Improvements in this field do not always require experimentation, for example in building algorithms, but ideas, like those in HCI, benefit from it. Is Computer Science not a science, or does it not have its own paradigms? While this is a debatable question, I think the fact that the paper does not defined what it thinks science to be is a glaring weakness.