March 2009 Archives

The Information System research is plagued by the ubiquitous problem of contexuality. That is, while studying any phenomenon, a researcher needs to ascertain the extent to which his findings are skewed towards the specific context his data drives from, and how much they can transcend and hold true across different contexts.

I put together a paper ( with Dr Sawyer) for iConference that seeks to shed light on practices of two dominant epistemological paradigms (positivist vs. interpretivist), and comes up with a model which incorporates their upsides regarding the contextuality problem. Although it is mainly oriented  toward Information Systems research, I believe its line of reasoning would be illustrative for other soft disciplines which wrestle with the same set of difficulties.
To download the paper, click here.

The following paragraphs have nothing to do with IST590, but I thought that different perspectives from philosophy of science might be illustrative when it comes to charting the boundaries of our field- I wrote this short essay for a course on Philosophy of Science :

Positivism:

The main objective of a positivist is to construct a solid foundation for the science (i.e. IST). This foundation should be free from metaphysics, and pseudo-concepts introduced by metaphysics. Metaphysics is composed of assumptions that cannot be reduced to logical simples, subsequently cannot be verified empirically. In this way, he lets the science and the empirical world speak for themselves. All prepositions are "pictures of reality." The method of enquiry is hence treated as an objective pursuit. By restring himself to strict measures, a positivist seeks to adopt a genuinely natural point of view. He is looking for certainty and the certainty is generated by going about empirically and check out everything against nature.

His view toward the empirical world and its significance would lead to a number of implications for the kind of science we are doing at IST. The science is taken as a consolidated body of propositions. In other words, there is a one science underlying of various scientific disciplines; and this cornerstone reflects one real world. According to this concept of unity of science, different sub-disciplines within IST could not be treated as separable while they are derivatives of the neutral elements of experience. After all, all these scientific disciplines contain theories or research traditions that are largely commensurable.  What the "I" people are doing is not inwardly different from what their "P" and "T" counterparts come up with. Essentially they are digging into the same ground but from divergent sides. Whatever they pull out is anchored in the same principles (empirically verifiable), and according to the precise and strict measures of positivisms these findings are leading us to the same truth which is out there in the natural world. To this end, the old library since is not different from IST. Although the names might bring up different connotation, they are multiple conduits to the same reality. 

The reductionism put forth by positivisms also makes us treat different phenomena in our field almost alike. For instance, the way we might look at a purely technological enquiry like the evaluation of search engine agility, could not be that different from investigations of the perception of the users of an eGovernment website. A positivist strives to reduce even social phenomena and look at them, for example, as relationships between a system competency and respective behaviors of its users.   Within IST, outwardly different phenomena like people and technology ought to be studied alike.  When it comes to people and the complexity and uncertainty around their actions, we reduce our model and simply dispense with those qualities that are not empirically verifiable. 

Positivism also suggests its own method of enquiry for our enterprise. Given the objectiveness of the truth, the method is dissociated from the personality and social position of the investigators. That is to say, the results engendered by IST researchers are substantially transcultural and context independent. In this sense some of the current approaches and research within IST could not be treated as valid attempts. The sort of analysis that Andrea generates out of a corpus of interview is somehow bound to her own interpretation. She is essentially implicated in the research and has a great bearing on the instrument of the research. As such her approach does not satisfy positivist sort of criteria where the objectivity of the research is the one of the main precondition of research.  Her assumptions are also at odds with the kind of propositions that cannot be reduced to empirical prepositions. These non-extensial propositions are eliminated by positivists. The firm framework of empiricism also annihilates some of the stuff that is currently studied within IST. For example, in some research the possible outcomes could be regarded endless (particularly on the P side). However a positivist cannot tolerate any assertions of infinite collections or magnitudes.

 

Popper:

Popper believes that science undergoes a rational evolution. He also puts forth that it is possible to arrive at a judgment as to which of existing theories is closer to the truth. When we are exposed to different theories, we would accept the one that has excess empirical content, explicating everything that was formerly explained by the other one, and brining about some predictions that have been confirmed by experiment. For Popper inductive reasoning is a myth. Observing positive instances and then believing that the more positive instances, the more confirmation of a theory is not acceptable for Popper.   To him, it is not plausible to assume that having more inductive evidence would support a theory compared to its rivals. That is why he suggests an asymmetry between verification and falsification (as his own alternative). As opposed to positivists, he claims that no hypothesis is ever confirmed. Instead he directed attentions to the concept of corroboration. A hypothesis is corroborated if it has not yet been refuted and has stood up sever tests (i.e. attempts to refutation). This to say that rather going about purely experimental judgments, as with positivist views, we accept a theory until it is falsified. But a theory or statement is scientific if an observation can assert that it is false.

According to Popper we have to look at what we are doing here as an evolutionary science and a series of theory rather than theories. We need to take into account the relationships between theories. Although we may find that a couple of experiments are at odds with a theory, we don't dismiss it until a new theory is found which tells us the same things as the old one but without the difficulties of the old one. Here the conflict does not take place between theories and experiments but between rival theories.  A theory is regarded falsified when it is superseded by a theory with higher corroborated content. So Popper shifts the problems of how to appraise a theory to the problem of how to appraise a series of theories. So a series of theory rather than an isolated theory are said to be scientific or unscientific. To this end, falsification of Popper does not simply consider the relation of a theory with the empirical world, but it puts emphasis on the growth of empirical content engendered out of the completion of series of theories. This conceptualization presumes a cumulative process which is achieved by an articulation or extension of old paradigms. What is currently being done in IST does not revolutionize previous paradigms like library science. In contrast it is building on them, and attempts to get closer to the truth through an incremental process.  

Popper believes that the purpose of science is moving toward truth. The evolutionary process discussed above is supposed to move on to theories with more lines to truth. So we are always on the way, and cannot touch truth (like what positivists claim). That is why we have to be open to criticism even when it comes to basic statements, making sure that we are moving forward. In this sense, every theory is valid for the time being, and no solid foundation could be assumed for science as we proceed with competing theories.   We cannot for sure contend that the approaches we are currently drawing on within IST are in full accord with the absolute truth. Rather, we must make sure that on the grounds of Popperian falsification we are on the way. As such, the process of scientific evolution is put weight on, rather than the sort of instant results (embracing the absolute and objective truth) positivists are looking for. This however does not imply that every sort of theory can work its way into our circle. We are only receptive to a theory if and only if it's falsifiable. Hence, we might need to get rid of some of the theories we are currently employing here at IST, particularly the ones which do not lend themselves to the Popper's demarcation criterion.

 

Kuhn:

Kuhn bases his discussion on a descriptive account of the trajectory of science. One of the pivotal concepts that functions as the touchstone of his argument is the notion of "paradigm." A paradigm could be characterized as a set of theories and the kinds of methods that they share. A scientific community is formed around the paradigm and the paradigm is faithfully pursed by the related scientific community. We can treat information science as a new paradigm. It embraces the whole network of theories, beliefs, values, methods, and educational objectives of a scientific community. According to Kuhn, it provides the bond of the community and shapes its world-view and guides its research. At a more concrete level, it advances a set of guides of actions (what is called rules by Kuhn) that members (i scholars and students) are supposed to use in their scientific endeavor, and further articulate the paradigm.

I believe the new paradigm is stemming from a series of older paradigms like computer science. Over the last decades or so, a number of anomalies have arisen that were not well-addressed by computing conceptualization. Phenomena like lack of user acceptance have plagued many IS implementations while computer science did not have enough handle to cope with these emerging challenges. The rise of these anomalies signifies a serious decline in the puzzle-solving efficacy of the old paradigms. As a result, the new paradigm of Information Science has been brought about by the scientific community. Kuhn believes that the new paradigm should now solve some outstanding and generally recognized problems that can be met in no other way. The IST paradigm hence needs to promise to preserve relatively large part of the problem-solving ability that has accursed to science through its predecessors.

However, this is not to say that the whole process of paradigm shift is rational in a sense that is preached by scholars like Popper. Kuhn does suggest that there are no algorithmic procedures to decide which theory fares better. Although competing paradigms are treated as incommensurable, they can be compared. The incommensurability clarifies what is involved when we compare alterative and rival paradigms. The computer science paradigm can be logically incompatible with the IST paradigm, incommensurable (cannot be measured against each other point by point), but comparable. That is, they are capable of being compared with one another in multiple ways without requiring the assumption that there is a fixed, universal measure which can hammer out both paradigms. As a result the shift in paradigm can be due to multifaceted, and multilevel changes that take place in the scientific community. The rise of IST paradigm could be due to the striking penetration of computing technology into organizations in 80s and 90s and particularly the internet boom in late 90s. As such, the scientific change - from one paradigm to another- cannot be governed by sole rules of reason.

The construct of incommensurability can have other implications for what we are doing at IST. Kuhn believes that the world changes as a new paradigm is adopted. This is to say that the proponents of competing paradigms practice their rituals in different worlds. In other words, the two groups of scientists see different things when they look from the same point of view in the same direction. A computer scientist investigating a phenomenon can come up with completely different results from what his information scientist counterpart has conjectured by looking at the same thing.

I think by interpreting Kuhn writing, one can conclude that a paradigm shift in our field has taken place. But IST as a paradigm is far to reach the normal science status. During normal science new findings are forced into performed and relatively inflexible boxes that a paradigm supplies. So no innovation is intended or expected. The paradigm is not also tested. Similarly no attempts are made to review models, values, commitments of a paradigm. I think the existence of a course like the philosophy of science which embarks on different epistemologies which are practiced in the field is a clear sign that we are have not arrived at such an epoch. We have not taken for granted things like methods of enquiry, values, assumption and so on.

  

 

Lakatos:

Lakatos aims to propose a rational reconstruction of growth of science. Therefore he advances the notion of research program. The research program is the unit by which he appraises the nature and evolution of science since he rejects the emphasis on individual theories. Rather he looks at the comparison of sequence of theories in different research programs. According to Lakatos, a research program includes a hard-core, the negative heuristic, and the positive heuristics. I believe IST can be thought of a budding research program because now we are able to work out all these three constituents for it. By considering IST as a research program, we would assume continuity in the kind science that we are doing here. This continuity and tenacity of the theories that we are currently employing can be explained if we look at our science as a battleground of research programs rather than of a host of isolated theories. In this way, we can detach ourselves from the specific choices and specific situations that we are in for the time being. Rather than trying to appraise our choice instantly, the unit that we examine is our research program. This lets us look retrospectively at our program which develops and changes over time and is consequently historical.  

Lakatos' criterion of demarcation is aimed at making distinction between mature science, consisting of research program, and immature science consisting of a mere patched up patterns of train and errors. I believe what are happing in venues like iConferance are attempts towards crystallizing disparate subfields into a mature science. By definition, mature sciences consist of research programs that anticipate novel facts as well as novel auxiliary theories, so unlike immature sciences, they possess a striking heuristic power. In iConferences, iScientist converge to consolidate the power heuristic of the filed which further would lead to the autonomy of our theoretical science.

IST as a research program must have developed a hard-core, the negative heuristic, and the positive heuristic. IST seems to move towards a momentum where all community members more or less would share a hard-core consisting of a set of theoretical hypothesis. These hypotheses would be looked at as refutable facts. They are also deemed to immune to revision. Then, it is the negative heuristic that would protect the hard-core of our field and would forbid us to attempt any revision. We must however articulate a set of auxiliary hypothesis which can function as a protective belt in the face of anomalies. This protective belt could be considered as the set of secondary hypotheses we are putting forth in our research. They have to get adjusted and readjusted or even completely replaced to defend the hard-core. For instance, different conceptualizations about the impact of technologies could be classified in these categories while things like technological determinism have been replaced by newer and more succinct frameworks in the face of deficiency. This process of articulating suggestions as to how the IST could develop, either in the face of anomalies, or as attempts to accommodate new phenomena constitutes the positive heuristic of IST research program. Separating positive heuristic and the hardcore let us bring about a revolution or creative shift in positive heuristic which can revive our research program when it is getting into a degenerating phase.

Lakatos also allows us to dispense with the naïve falsification measure which forces us to refute a theory and replace it in the face of counterexamples. For example if a phenomenon like Youtube emerges and cannot be accounted for by our research program for the time being, we need no refutation to tell us that our program is in urgent need of replacement; in fact the positive heuristic of the program drives us forward anyway. Lakatos believes that such an instant rationality and learning is bound to fail, and rationality works much slower that most people tend to think. So in this way we permit our research program to develop, to show its hidden strength and we will judge it only in the long term. Our measures only are applied "with hindsight", after the occurrence of either "degenerating" or of "progressive" problem shifts.

 

 

Feyerabend:

Feyerabend advocates a view which does not assume the development of science to be entirely rational. What is central to Feyerabend's construct of science is overthrow of any universal and overarching standard of demarcation.  He does not see any such universalistic rule which can assess claims of competing paradigms. Therefore, according to him methodological rules generally do not contribute to scientific success.

He further embarks on the principle of tenacity which let a theory survive in the face of counterexamples. In a sense he relaxes the strict reliance on the empirical evidence. On this ground, we cannot come up with a theory that could be completely consistent with all relevant facts. This would root out naïve falsification which asserts that scientific theories should be annihilated if they don't agree with known facts.  I have seen lots of self-righteous claims in our field where their preachers strive to abolish contesting agendas, arguing theirs are more compatible with the truth. Essentially Feyerabend is critical of these sorts of claims and any guideline that aim to judge the quality of scientific ideas by comparing them to the known facts.

Feyerabend advances his thesis by complementing the concept of tenacity with proliferation construct. He attempts to establish an interplay between the two. We might drift to inertia by solely emphasizing tenacity of scientific theories. He thinks we should be able to introduce new ideas even if the current theories are well justified and well-entrenched.  IST could be regarded as an exemplar of proliferation of diverse and even conflicting paradigms. According to the construct of proliferation, paradigms like psychology, sociology, computer science, and etc can live together although they might seem incongruous.  However, upon the arrival of a newcomer, they have to open some room and "breathing space" for it. There is no need to suppress even the most outlandish sort of theory which tries to work its way into IST.

To further his thesis, Feyerabend also puts forth the incommensurability of scientific paradigms. The standards of rationality of each paradigm are radically unique. So there is no general universal measure which can evaluate these incommensurable paradigms; no practical comparison between these theories could be made.   In other words, we are not able to work out any measure to evaluate as to what theories would get us closer to the truth. What Lee Giles and Andrea Tapia are doing after all is their own creations, which include their own standards they impose upon IST. Their theories and views could be empirically disconnected. So the choice between their theories would become a matter of taste. Hence, no obligation could be imposed upon the researchers within the field to go for one of them since we cannot claim for sure what approach reflects a fuller picture of the truth. Incommensurability, as defined by Feyerabend, implies that switching from an old paradigm to a newer one always implicates some gains and losses. This view encourages coexistence of competing paradigms within our field insofar as each provides us with things that are absent in the other.

The Feyerabend's view can have some overt and covert implications for the sort of direction we should move forward. Students of the field should not be prevented from speculating along different lines of enquiry. This is to say no reduction should be made in the number of accepted and comprehensive theories within the IST community. Feyerabend strongly argues against closing our mind where hard-wiring to established paradigms and practices diminish the tendency to explore unknowns. Feyerabend would also turn our attentions from our sense of community to our individual creativity.  In his view some of the collective mechanism like iConferance, presence of different concrete angles (I,P,T) which are aimed at consolidating the field could be counterproductive. By these mechanisms, individual members of the community might be pushed toward conformity to certain accepted practices. He believes every community even the ones considered to be open-minded, and enlightened are in grave danger of becoming closed, rigid, and indolent to new ideas and alternatives. In short, Feyerabend would argue for a pluralistic methodology which sees the critical power of science in overthrow of fixed orthodoxies.

 

Hermeneutics

Hermeneutics has manifested itself as a robust alternative to Scienticism. It has essentially emerged as a defensive reaction against universalistic and reductionist claims made in the name of science. These claims aim to assimilate whatever sounds legitimate, acceptable and rational in their view into knowledge, and throw away whatever that cannot be reduced to cannons of scientific discourse. Hermeneutics would preclude us from adopting any such measures for falsifying a variety of science and paradigms we are practicing here, at IST. We cannot come up with a single silver-bullet-like demarcation criterion which judges what is strange and alien to us as if this was the sole and exhaustive measure of rationality for the whole spectrum of our field. If each distinct sub-field that constitutes our community views its alien counterparts in the above fashion, we could end up with disconnected islands which are attached to their orthodoxies (whether rooted in sociology or statistics), possessing tons of misconceptions about one another.

Hermeneutics portrays a distinct picture of truth and knowledge. It does make no distinction between reason and tradition. Reason gains its distinctiveness power always within accompanying traditions. Different traditions within IST bear their own rationality. These rationalities could not be accounted for unless we attend to the very historical contexts they are anchored in. For instance, one cannot appeal to the contingencies of the "P" people without delving into the sociological, psychological, and philosophical bedrock of their horizon.

The concept of incommensurability is central to Hermeneutics where no single language is thought as appropriately explaining contesting theories.  IST is no exception to this, hence consists of incommensurable and at times incompatible paradigms. However, this is not to say that we are by no means able to compare them and flesh out a sort of collective rationality. Although there does not exist a fixed grid by which we can hammer them out, Hermeneutics contends that they need to engage in a dialogue in order to arrive at a rational ground. As mentioned before, different traditions within our community bear their inevitable prejudices. Our prejudices however are not necessarily unjustified and erroneous, so that they inexorably distort the truth. These are simply conditions whereby we experience something. I, as a student of Steve Sawyer and a person loyal to the pillars of social informatics, am not able to bracket all my prejudice and background while I am engaged in this dialogue. But the sort of dialogue sanctioned by Hermeneutics requires us to avoid imposing our beliefs, and classifications that are well entrenched in our tradition. This precondition for fusion of horizon lets us understand ourselves better and make adjustment based on the interactions we have had with the alien traditions.

I believe an exemplar of this sort of discourse is the one into which Rob Kling (as a computer scientist on one side) , and Steve Woolgar and Keith Grint (on the other side as sociologist) did get (See [1] [2] [3] [4] [5] for more details on the series of discussion). They immersed themselves in a discourse to see how the material properties of technology situate in a conceptualization of technological agency. While appealing to the other's horizon, they did not retreat from their own tradition.  I believe this could be a case to which other cross-disciplinary discourses in our field could look at as an exemplar. This collective rationality however was reached by becoming more aware of their own blind prejudices (i.e. what cannot be explained by sociology and what cannot be explained by computer science models), learning there is more to the "truth"  then is captured by their own entrenched view of the world. Here is where Hermeneutics distances itself from relativism when together we can enlarge and enrich our horizon by understanding alien horizons. Basically fusion of horizon brings us above our own horizon.

In conclusion, I think we are still in the quest of a kind of communal rationality insofar as we have not become a well-entrenched field. Our identity and scholarship are not as consolidated as those of business schools; even each iSchool opted for a different name. Our scholars have by and large been trained in other colleges, so attending different conferences ranging from computer science to sociology.  We are more prone to draw on Hermeneutics' agenda and interact with other paradigms. But there is a possibility that after a while we become like computer science departments or business schools that once upon time used to enjoy the same caliber that we reflect today. At that time, we might be somehow close to other paradigms when we would have our solid identity, our conferences, and when our students would be taught by professor who have obtained their PhD from iSchool and would have been indoctrinated within "iParadigm".

 

 

1.            Grint, K. and Woolgar, S. Computers, guns, and roses: What's social about being shot. Science, Technology and Human Values, 17 (3). 366-380.

2.            Kling, R. Audiences, narratives, and human values in social studies of technology. Science, Technology and Human Values, 17 (3). 349-365.

3.            Kling, R. When Gunfire Shatters Bone: Reducing Sociotechnical Systems to Social Relationships. Science, Technology & Human Values, 17 (3). 381.

4.            Woolgar, S. The Turn to Technology in Social Studies of Science. Science, Technology & Human Values, 16 (1). 20.

5.            Woolgar, S. and Grint, K. Computers and the transformation of social analysis. Science, Technology, & Human Values, 16 (3). 368-378.