• View
• Edit
• History
• Print

Paradigms

Questions

• What relevance do paradigms have for ways of thinking outside of organized science?
• What paradigms am I trained in, at what levels?
• What paradigms shape my approach to computers and cognition?

Assigned Reading

• Winograd and Flores, Our Path (Chapter 1).
  Readings from Winograd and Flores are from the textbook, Understanding Computers and Cognition, Addison-Wesley, 1987.
• Thomas Kuhn, The Structure of Scientific Revolutions, Chicago: University of Chicago Press, 1962.
  • Chapter II: The Route to Normal Science, 10-22
  • Chapter VI: Anomaly and the Emergence of Scientific Discoveries, 52-65.

Additional Resources and Readings

• Background on Kuhn
• A short excerpt of Lakatos' article from Imre Lakatos and Alan Musgrave (eds.), Criticism and the Growth of Knowledge, Cambridge Univ. Press, 1970.
• Donald Stokes, Pasteur's Quadrant

Responses

Do the social sciences operate under paradigms? - Chris Anderson

The short answer to the question posed above seems to be no – but I want it to be yes! I am drawn to a particular passage in Section II of Kuhn's The Structure of Scientific Revolutions. He writes:

The introduction of Understanding Computers and Cognition lays out an argument for changing the way one thinks about cognition. It also, however, builds up that argument from its basic concepts, without referring to textbook ideas. There are, of course, textbooks about computers, cognition, rational thought, and so on – but Understanding chooses not to use them. Does this mean the field to which Understanding is written does not yet have a thought paradigm? Furthermore, what field are Winograd and Flores addressing?

The back cover of the book mentions professionals from design and from artificial intelligence. Do those fields yet operate under a paradigm? Assuming "design" means "computer design", it may be safe to say that no, the field does not operate under a scientific paradigm. My intuition is that most works of design are free-standing and do not rely on a previously established body of work. Artificial intelligence, on the other hand, seems different. Since it is part of the study of computer science, it should have a paradigm. However, we are concerned in Understanding about questions from a more distant perspective than the one that computer scientists tend to look at artificial intelligence with. Perhaps this field does not have paradigms – perhaps it is more social in nature.

I am not convinced that the social sciences do not operate under any sort of paradigm. Their work is often more speculative than "harder" sciences, but it nevertheless builds off of previous bodies of work. Their professionals discuss with one another the results of their research via journals (not necessarily books). One key feature that is only debatably present is "knowing with precision what [to] expect" (Kuhn 65). How do I go about doing an experiment in a social science like design? I am not sure what my metrics are.

We can probably think of better examples of social sciences with paradigms than computer design. Philosophy seems like a strong candidate. The work of Saul Kripke has caused much of a restructuring of the philosophy of language in the 20th century. Kripke built upon his predecessors' work, and always argued inside the philosophical framework. Does this qualify his work as causing a paradigm shift? I would like to discuss the relationship of social sciences and paradigms more, though I do suspect the later parts of Understanding deal with this issue to some extent.

Ramblings about Paradigms -Vala Dormiani

My quibble about the idea of Paradigms is that Kuhn presents it quite rigidly. I interpreted a paradigm to be the creation of concrete models that form the basis of a field and act as the fundamentals. Subsequent scientists in that field would build on the shoulders of others that have formed this paradigm and would no longer need to reinvent the wheel. Kuhn calls this the “genesis and continuation of a particular research tradition”. My issue is that Kuhn often presents the creation of a paradigm as one revolutionary work such as Franklin's “Electricity” or Newton's “Opticks”. These scientists built on the shoulders of their predecessors to some degree. Why are those prior works not seen as paradigms? Granted, Kuhn makes the point that the “transitions have seldom been so sudden or so unequivocal as my ...discussion may have implied” so he is acknowledging this potential issue. Kuhn mentions that “one field of study after another has crossed the divide between...its prehistory as a science and its history proper”. This crossing is the paradigm idea. However, I would argue that the moment of crossing is ambiguous. Similarly, the protagonist who is often credited with pushing a science across this thresh-hold is up for debate.

Are paradigms relevant to aspects of life that are not the normal sciences? I took the normal sciences to be techie sciences. Paradigms are obviously relevant to techie sciences. For example, an electrical engineer does not constantly reinvent the computer. He/She attempts to make advances to the existing basic structure. Similarly, as a CS major, I don't constantly create new programming languages to write code. I simply build on the shoulders of other and use existing languages. However, does the paradigm idea apply to the social sciences? I agree with Chris that the short answer is no. In the social sciences, the book is still seen as “a vehicle for research communication”. The layman can potentially follow the progress in the field as the professionalism and specialization has not been solidly established and esoteric vocabulary and skills are yet to be developed. Kuhn characterizes this as pre-paradigm. In my opinion, a field such as history can definitely be characterized as pre-paridigm. The book is still the main method of conveying research. The conveying of past facts is generally not a pursuit that builds on the shoulders of others and is stand-alone. However, as Chris points out, philosophy might be a different story. Also, is a pre-paridigm field an inferior field? I say no.

Feelings -nathan

Are computers just brittle replicas of real life "thinking"? I would say that complexity theory (with some assumptions) tells us that computers can solve all solvable problems, therefore they can solve all problems that humans can, its just a matter of how fast, and by what method. Is there anything that humans do besides solve problems? Clark raises the point that we do seem to experience feelings and poses this question:

"Does emotion depend on meat or can we build it into any material?" I like Clarks thought experiment where he talks about a person having their brain slowly replaced with silicon that mimics the same behavior as the original meat brain. The assumption is that the meat brain and the silicon brain are going to have the same exact input and output, thus the behavior of the human should be the same with either brain. But then the conclusion that Clark draws is that as the meat brain is replaced the human would steadily start to feel like a zombie. The assumption tells us that the behavior would still be the same so even if the human lost all feeling, if we asked that human "hey man, how do you feel about this pizza?" He or she would still respond, "it tastes great," even if they "felt" no taste at all! Similarly, "are you happy about your new car" they would reply just as they would with their meat brain, "oh yes, very excited." The human would even smile and act just as though they were experiencing those emotions, but the emotions would all be simulated. How can we be sure that we are feeling emotions ourselves? If we can build machines that mimic perfectly our own input/output then how do we know we are feeling emotions and not just saying we are? I suspect as we get closer and closer to building machines capable of passing the "Turing test" we will learn the answer to this question. Is there something in the meat that makes us feel or is it purely in the instructions, perhaps some combination?

As a side note about paradigms I thought it was funny how Kuhn mentions that if no one believes in your paradigm you find yourself in the philosophy department : )

---nathan

Thoughts regarding my own interdisciplinary field after reading Kuhn - Luke Dahl

On beginning to read Thomas Kuhn’s ‘The Structure of Scientific Revolutions’, the first things that struck me were the power of the idea he presents and the clarity of his presentation. I also found it refreshing and liberating to consider the social and historical dimensions of science. Too often, it seems to me, the dialog about science proceeds as if science were some inevitable force and we people are but the means by which it progresses. But in actuality science is a human activity, and our social practices as well as our own psychological tendencies are the environment in which it takes place. On the other hand, Kuhn’s description of the process by which new paradigms emerge does have a slightly fatalistic component: we cannot predict or deliberately create a paradigm shift. It occurs only when the right conditions have accidentally coincided. But that’s not the main point I wish to discuss...

The second thing that struck me as I read this article was a sense of dread. I realized that every description of pre-paradigm science (or whatever we would like to call the activities that precede the first paradigm), was a description of my own field. My field which I call “Music Technology”, but which is also sometimes known as “Computer Music”, is interdisciplinary. It includes work in music composition, music theory, musicology, music cognition and perception, computer science, human-machine interaction, digital signal processing (in sound synthesis, processing, analysis), sound recording, information retrieval, and I’m sure a few others I’m forgetting.

This interdisciplinarity creates the conditions Kuhn uses to describe the pre-paradigm science: no commonly agreed upon body of work to take for granted, constant reiteration of fundamentals, a lack of standardized methods, and research that seems often to be haphazard and unfocused. Perhaps Kuhn had no intention of describing interdisciplinary work. He excludes from the sciences crafts (including medicine!) that serve a social need, and he makes no mention whatsoever of the arts. However the similarities are striking to me.

It is interesting to compare the role of novelty in science and in (at least my own) interdisciplinary work. Kuhn points out that normal science does not strive for novelty. However the standardization that accompanies a paradigm allows the astute practitioner to perceive anomalies, i.e. deviations from the socially agreed upon expectation, which, when explored, may precipitate a shift of paradigm. In my interdisciplinary field novelty is prevalent, and often seems to be the goal. Perhaps this is due to the artistic component of my field. Or perhaps it is because, lacking a consensual single research goal, novelty is the only quality that allows work to stand out.

Two other questions I have, but have not yet pursued are: 1) Is it possibly for a first paradigm (and therefore a science) to emerge from interdisciplinary work? (And also would I really want to work in that new field?!) And 2) How do people evaluate the quality of interdisciplinary work, when they lack the standards that come with a paradigm?

Strong Opinions, Weakly Held - MFSchar

Is 'revolution' continuous with historical tradition? - Jeff Wear

Of the development of new fields of scientific inquiry, Kuhn writes in The Structure of Scientific Revolutions that those fields'

But for the most part, Kuhn describes scientific progress in terms of a gradual continuum: paradigm change is preceded by a a "gradual and simultaneous" "emergence" of discovery and invention of fact and theory (62). As such, "revolution" may not be an appropriate description of paradigm change.

Kuhn dismisses the idea that a discovery can be localized to one instant in time. The initial discovery ("novelty of fact") must be confirmed by subsequent experiments and given meaning and context within a (new) theoretical framework before a paradigm shift can say to have taken place. "If observation and conceptualization, fact and assimilation to theory, are inseparably linked in discovery, then discovery is a process and must take time" "the progress of normal science... prepared the way to a breakthrough." (53). At one point does one paradigm stop and another begin? Perhaps it is even possible to historicize science, to demarcate paradigms, in many different ways according to the historian's whim.

In any case, it is unlikely that scientific "revolutions" have a seminal moment, an instant in which the old order is overthrown as it might be in a political revolution. But to what extent is the old order even overthrown at all? New theories do not necessarily reject the observations, equipment, etc. of the old, and maybe only some of the explanations. New theories do not represent complete breaks with the tradition that has come before, but instead may extend those theories; certainly they must account for the same phenomena. It may even be that the old ways of thought are sufficient and necessary to prime the investigator for discovery and thus for paradigmatic change. Kuhn notes that "anomaly appears only against the background provided by the paradigm"; normal science, which enriches itself with more elaborate equipment, skills, etc. increasingly provides an "indicator of anomaly and thus an occasion for paradigm change" (65). Paradigm change does not occur in isolation and could not occur without the benefit of the scientific tradition that preceded it.

Given the connotations of "revolution" - a term which never appears in the actual text itself, if I am not mistaken - Kuhn appears to be describing less the "overthrow" of a technosocial order and more simply the nature of scientific progress. In a historical sense, it is possible to mark how traditions have developed over time and to mark great contrasts at the endpoints of an interval of discovery, but I think that the curve connecting those points is more smooth than broken, its ascension marked by steeper and flatter periods rather than jumps. Perhaps this is an instance in which commercialism, not truth to content, determined the title of a text.

After-Class Comments

In class, we talked not only about the development of paradigms as espoused by Kuhn, but also the links between scientific progress and agents such as patrons, beneficiaries, supporters, and amateur contributors - agents that must be regarded as being located strictly outside the paradigmatic loop. My ongoing interest in the real-world grounding of the loop leads me to ask whether the idea that mature science is or can be defined by a single cycle might be a bit restrictive. I wish that Kuhn had discussed some of the obstacles to the revolution of the scientific cycle that even "mature" fields face.

I'll frame the scientific cycle's foundation in an ongoing search for truth - for a more perfect understanding and ability to explain natural phenomena. We discussed the difficulties that the social sciences might encounter in operating a paradigm in the face of competing schools of thought and few metrics to define 'truth' or in any case to rally the field's practitioners around one theory, but this confusion of theories is not limited to the social sciences. Physics nowadays, for instance, is split between many abstract theories such as string theory and brane theory which cannot, for the foreseeable future, be confirmed by experimentation. Without experiments to confirm the predictive power and thus 'truth' of theories, where does this leave science? Can several paradigms operate in tandem to produce valid scientific knowledge? My instinct is that the wheel cannot break loose of its moorings and still be said to produce scientific knowledge. My English teacher last Spring cautioned against easily forming a straightforward, definitive argument. She preferred complications and a nuanced interpretation. But I feel as though even in the face of extreme difficulty (such as the possibly flawed rationalistic assumptions we are challenging in this course), science must strive for something more empirical, something which is arguably "true". Science ultimately attempts to describe how the world is, not how it might be. Without experimental confirmation, there is nothing to recommend one theory over another except intuitive appeals to elegance, at which point science has become philosophy or art. -- Jeff Wear