- The General Approach
1.1 Philosophy, Psychology, and Artificial Intelligence
Paul Thagard’s response to the critiques of his article “Explanatory Coherence” begins with a discussion of the criticisms that Dietrich and Wetherick pose about categorizing Thagard’s theory of explanatory coherence (TEC). Dietrich mentions his confusion in distinguishing between whether TEC is a theory of philosophy of science or of psychology. Thagard explains that ideally TEC belongs in both, but this may not always work out perfectly. Dietrich concludes that it makes slightly more sense for TEC to be a psychological theory because he does not see a point to ECHO if TEC is a theory of philosophy of science. Thagard clarifies that the point of ECHO is to provide a more in-depth example of TEC than has previously been supplied. In addition, Thagard notes “TEC and ECHO…are not positivist because the emphasis in on high-level theories, not on observation, and data can be rejected; and the principles of explanatory coherence go well beyond formal logic” (491). Thagard coins the term “biscriptive”, a combination of descriptive and prescriptive, which describes how people make decisions when in accord with the best options available, therefore offering insight as well as a critique into how human performance functions.
Wetherick does not think that one theory can be applicable to both psychological processes as well as to sociological ones, and that TEC is sociological. Thagard recalls his explicit use of Lavoisier and Darwin as models for ECHO, and that his “examples come from the history of science, not its sociology. Nor do I pretend to model minds free of prejudice and preconceptions” (491). This is also a response to Wetherick questioning the success of a connectionist model when emulating a prejudiced mind. Wetherick argues that conscious symbolic processing always determines perceived explanatory coherence, and that Thagard does not include symbolic processing at all in his model. Thagard responds by mentioning that people often find coherence after they have stopped consciously thinking about it, and therefore explanatory coherence is not always determined by symbolic processing.
1.2 Connectionism
Thagard acknowledges that the ECHO relies mainly on connectionism to demonstrate TEC, but Dietrich believes that ECHO is not directly related to TEC. Thagard argues that ECHO is vital to TEC because it offers examples to back up this theory and that TEC evolved from the creation of ECHO. Lycan questions Thagard’s heavy use of connectionism by arguing that connectionist architecture has no advantage over von Neumann architecture and that ECHO’s use of connectionism is actually not a strong trait in the model because it could achieve the same results by using a traditional architecture. Thagard responds that connectionist models act as supplements to traditional architecture. A nonconnectionist version of ECHO can be used, but the connectionist model results in more appropriate conclusions.
- Theoretical Issues
2.1 Explanation and hypothesis evaluation
How do we define explanation? Achinstein wonders if we can use Thagard’s theory successfully without also having an operating definition of explanation, to which Thagard expresses his current incapacity in creating a theory that incorporates all types of explanations. Achinstein also describes how there needs to be a connection between explanation and acceptability. Thagard retaliates how “the inference from ‘H is the best explanation of the evidence’ to ‘H is acceptable’ does not require any special relation between explanation and acceptability” (492).
O’Rorke brings up the interesting issue of how an agent’s goals and priorities are important factors in evaluation, yet Thagard thinks that these factors only matter for the generation of hypotheses and not whether an agent will accept a hypothesis or not. Thagard mentions his desire to create MOTIV-ECHO, a program that will “reach conclusions on the basis of how well beliefs satisfy its goals, as well as on the basis of how much explanatory coherence they have” (493). Thagard also describes how people tend to supplement their own desired conclusions with evidence that they have selectively applied to a situation, and will ignore evidence for an unfavorable conclusion. On a similar note, Sintonen recognizes that people often make choices because of the promise they show, not necessarily because they have coherence. Thagard agrees and adds that because of this, people do not always choose the best available option. In developing MOTIV-ECHO, Thagard hopes to incorporate this phenomenon as a part of the program’s ability to rationalize.
2.2 Simplicity
Reggia emphasizes and exploits the claims of simplicity, which Thagard defends, to suggest the alternative use of Bayesian methodology. Thagard confidently responds that this interpreted ‘principle of cardinality,’ where “explanatory hypotheses with the smallest number of hypothesized components are preferred” (p. 493), is not always the case and instead has more so to do with the configuration of hypotheses. Furthermore, he states that frequency and probability-based hypotheses are not always found in the scientific and legal domains to which ECHO is applied, and therefore may not function as the best alternative for TEC.
2.3 Analogy
While Thagard believes that analogies may improve the coherence and acceptability of hypotheses and theories, McCauley disagrees by saying that theories have a role in determining the nature of analogies and thus cannot be evaluated in this way. Similarly, Hobbs argues that these analogies improve the explanatory power of theories that rest on abstract principles. Such theories and abstract principles, however, do not have any influence over analogy recognition, which according to Thagard & Holyoak, can be achieved through the examination of pragmatic, semantic, and structural constraints. Although ECHO only addresses pragmatic constraints, many other theories (i.e. natural selection) appear to address individual constraints depending on their intended use of analogies. As for analogical reasoning, Gabrys & Lesgold point out the distinction between the case-based reasoning of juries and judges. Despite their intentions, this critique offers no objection to the constraint-satisfaction model, which we find in law and computational programs, like ECHO and the analogy program, ACME.
2.4 Conceptual change
Many critics, like Giere, attack Thagard and TEC for its lacking intentions to adjust and account for historical change in science. Giere questions whether Thagard’s model truly represents the reasoning of scientists and whether it can explain the transition from old to new theories. Thagard does not provide a clear answer to these limitations, but hopes to further examine the breadth of ‘judgmental’ and ‘representational’ mechanisms used by practicing scientists to acquire understanding of explanatory relations over time until new theories form. Nevertheless, Thagard believes that psychological experiments are of utmost importance in understanding and representing human reasoning.
Mangan & Palmer, in contrast, adhere to the Kuhnian idea that scientific revolution changes the methodological principles of explanatory coherence faster than TEC can account for them. Thagard argues that these philosophers grossly exaggerated the historical variability of such principles. To support their argument, Mangan & Palmer describe Darwinian revolution as a shift in the methodological use of analogy to improve theoretical explanatations. Thagard, however, finds evidence of this shift in the writing of philosophers and scientists (Paley, Huvgens, & Fresnel) well before Darwin’s time thus contradicting the Kuhnian dogma.
2.5 Logic & Probability
Other critics, like Feldman, begin to challenge the semantic foundation of TEC in comparison to probability theory. Although examples of “clean and well-understood formal semantics” may be hard to find and incorporate into ECHO, logic and probability theory should have no greater success (p. 495). Thagard points out that the Tarskian semantics of predicate calculus avoid the ‘central semantic question’ by repeatedly assigning definitions to progressively complex formulas. Furthermore, probability is neither fully understood nor interpreted in terms of its definition and theoretical axioms.
Cohen emphasizes the importance of predictions, which when successful, may signal simple explanations to hypotheses without post-hoc additions to the original theory. He also criticizes TEC and ECHO for having no way to “determine the acceptability of a conjunction based on the acceptability of the conjuncts” (p. 495). While Thagard admits this to be true, he reminds us that probability theory fails to a similar degree in calculating conjunctive probability from the indeterminate level of dependence between conjuncts. Such probabilistic knowledge is rarely available to the domains of ECHO. Despite these quantitative limitations, Dawes uncovers ECHO’s ability to deal with the conjunctive complexity of Simpson’s paradox where supporting evidence may refute a hypothesis if both taken under consideration.
- Problems with the ECHO Model
Many commentators expressed problems with the ECHO model and suggestions for improvement. A recurring issue is the arbitrariness of inputs to ECHO when considering explanations versus analogies or hypotheses versus evidence. McCauley questions how disputants can neutrally decide what constitutes an analogy because virtually anything can be analogous. Thagard claims this skepticism is only relevant if the Kuhnian view, which implies that fundamental principles are malleable, is correct. However, to him, this view is exaggerated. Additionally, both Dietrich and Zytkow question ECHO’s ability to distinguish between evidence and hypotheses, but Thagard doesn’t find the distinction problematic either.
Bareiter & Scardamalia notice the problem that when competing hypotheses are inputted into ECHO, they can affect the weight and activation levels of each other, and therefore result in differing outputs. Thagard explains two ways to address this: enrich the input to notice contradictions that have already been omitted, or adjust the output threshold because once a unit hits subthreshold, it no longer affects the activation of others. Meanwhile, McDermott questions ECHO’s output in general, stating that it may be established from the weight of links between propositions based purely on the problem’s structure instead of its content. Thagard dismisses this as peripheral and states that the model as a whole shows how explanatory hypotheses can be examined in complex ways.
Finally, commentators present alternate methods to compare to ECHO. Hobbs describes a “naïve method” of counting propositions which subtracts the number of hypotheses a theory uses from the number of evidence pieces it explains (#E – #H). This method, Hobbs claims, arrives at the same conclusion as ECHO, so why bother using ECHO at all? Thagard then claims that there are numerous possibilities in which ECHO and the naïve method don’t yield the same conclusion, and therefore are not equivalent. Simon and Zytkow introduce the connectionist model STAHL and compare its abilities to ECHO. Thagard acknowledges STAHL’s abilities, but doesn’t agree that it is a more comprehensive model of theory evaluation. STAHL is a discovery program and therefore more effective than ECHO in considering proposition content, while ECHO models evaluation, so is superior to STAHL in that it’s not restricted to a single method of hypothesis evaluation. Thagard eventually acknowledges that collaboration of models could produce an improved method of theory evaluation.
- Psychological Adequacy
Many commentators question the psychological adequacy of TEC and ECHO. First, as Klayman & Hogarth state, the examples utilized in the target article to exhibit the use of ECHO in TEC are insufficient tests of psychological validity. They state that ECHO doesn’t model the process of thinking, but rather its end result, and therefore is incapable of representing cognition (478). Thagard asserts that researcher have begun launching experiments to grasp the empirical side of ECHO and support its psychological adequacy. Yet many authors, specifically Earle, express concerns about the testability of ECHO due to the subjective nature of what is considered rational when credible hypotheses conflict.
Cheng & Keane state Thagard’s account is too holistic and parallel to be deemed psychologically adequate. Individuals approach theory evaluation in a more piecemeal fashion than ECHO, which considers all hypotheses and evidence of a theory simultaneously. Thagard asserts that although this may not be within human capacity due to limitations in short term memory, most evaluations of the explanatory coherence of propositions happen unconsciously, and therefore represent parallel and holistic judgment. Therefore, although a cognitive model would integrate ECHO with conscious deliberation, Thagard believes this integration can occur.
Finally, several commentators (Chi, Bereiter & Scardamalia, and Read & Miller) suggest the use of ECHO to explore certain psychological phenomena, including conceptual changes within individuals, whether people can learn to evaluate hypotheses more effectively (and to encode pieces of evidence they disagree with), and the application of TEC to social phenomena. Thagard is enthusiastic about these approaches, yet notes that more research is needed in psychology, philosophy, and AI in order to push forward.
Discussion Questions:
- Is a working definition of “explanation” needed in order to efficiently evaluate TEC?
- Should we use a program like ECHO to ‘teach’ people to evaluate hypotheses more effectively in a rational way? What would be the implications of this in society, science, and law?
Considering my last comment, I would propose the following questions on Thagard’s Theory: 1) Which has greater limitations in processing and explanation: ECHO or the human mind? That is, ECHO has the advantage of handling and processing large quantities of evidence and information as a computer program not bound by human limitation, but are their uniquely human advantages to information processing and explanation?
2) Should ECHO be treated strictly as a theory of science, or is their room for it as a theory of philosophy or a theory of psychology, contrary to what Cheng and Keane propose?
I also apologize for the delayed response. A few of you have so far addressed the appropriateness of addressing Thagard’s theory of Explanatory Coherence as a psychological theory, and as a sort of medium between scientific theories and philosophical theories. I find that one of the most compelling arguments against this proposition, found in the compilation of criticisms against Thagard’s theory beginning on page 469, is Cheng and Keane’s censure of explanatory coherence a psychological model. Interpreting explanatory coherence as a psychological model, they argue, poses two primary problems: the first is that, because of the nature of human psychology, any psychological theory must acknowledge the limitations of human processing, which Thagard does not adequately do.
The unlimited capabilities of ECHO, they argue, are not compatible with a theory that draws much of its foundation from the limitations of the human mind. From a psychological perspective, they argue, humans would be better apt to handle a collection of explanatory evidence and propositions in a gradual, assimilatory form rather than a sudden, simultaneous introduction of new concepts and a large amount of evidence, as Thagard suggests. Cheng and Keane furthermore mention that this theory of explanation has been directly supported by “new experimentalist” research, which contradicts Thagard’s “holism” theory in the philosophy of science.
Secondly, Cheng and Keane argue that, because explanatory coherence does not specify the “psychological processes that underlie explanation”, there is no effective difference between asserting that a piece of evidence “explains” a proposition and asserting that it is simply “connected to it”. Because of these limitations, Cheng and Keane argue that ECHO should be confined to the realm of scientific theories rather than philosophical theories or in the ‘happy medium’ of psychological theories.
I also apologize for the delayed response. A few of you have so far addressed the appropriateness of addressing Thagard’s theory of Explanatory Coherence as a psychological theory, and as a sort of medium between scientific theories and philosophical theories. I find that one of the most compelling arguments against this proposition, found in the compilation of criticisms against Thagard’s theory beginning on page 469, is Cheng and Keane’s censure of explanatory coherence a psychological model. Interpreting explanatory coherence as a psychological model, they argue, poses two primary problems: the first is that, because of the nature of human psychology, any psychological theory must acknowledge the limitations of human processing, which Thagard does not adequately do. The unlimited capabilities of ECHO, they argue, are not compatible with a theory that draws much of its foundation from the limitations of the human mind. From a psychological perspective, they argue, humans would be better apt to handle a collection of explanatory evidence and propositions in a gradual, assimilatory form rather than a sudden, simultaneous introduction of new concepts and a large amount of evidence, as Thagard suggests. Cheng and Keane furthermore mention that this theory of explanation has been directly supported by the “new experimentalist” research of Ackerman, Franklin, Galison and Hacking
One of the debates present in the response articles is whether or not TEC is a philosophy of science or a psychological theory. Thagard later explains that he wants to find a medium between philosophy of science and sociology (or at least that is how I understood his explanation), and is psychology not a medium between science and sociology?
To me is seems that Thagard’s theory of explanatory coherence, laid out with the help of ECHO is more of a philosophical science based theory, rather than one that falls under both categories as is his ideal. While science can explain much of the way that humans think, I still find issue with science’s (and ECHOs) ability to take into account the un-explainable aspects of human reasoning and decisions that are driven by emotion rather than reason. This connects back to the definition of “explanation” and whether or not, for example, feeling an emotion towards a particular action is an explanation for a decision. And often we also alter explanations to fit our beliefs, this leads me to the question of what does Thagard believe happens when explanations don’t fit with out basic? What makes some explanations more believable than others? Are explanations ever objective? How can they not be influenced by experience? Thagard says that he is not pretending “to model minds free of prejudice and preconceptions” (491) but I struggle with the idea that he can accurately portray all the biases we hold as humans and how they interact with our decision making skills.
Another major problem that I have withe Thagard that was pointed out by Reggia is the idea that simplicity should be upheld as one of the best criteria for sustaining a hypothesis. Just because something is implausible does not mean it is not possible and simplicity does not always yield the best explanation.
Thagard claims that his theory of explanatory coherence can be used in psychology, philosophy of science, and in legal reasoning. Reading the first half of this article, I wasn’t, and still am not, convinced that his theory could be used in psychology and represent human reasoning. I mentioned in the last response that ECHO cannot realistically claim to represent human reasoning because everyone has biases and a different weights are paced on different observations or pieces of evidence. I thought that ECHO might be a better way of reasoning because I believed that it would remove these biases. However, Earle mentions that “different end states may be reached, depending on the initial settings of the parameters and on the priority given to a certain piece of data” (474). Does this mean that ECHO itself also has certain biases depending on the initial conditions inputted by the programmer? How do we decide what to include in the initial conditions and what to omit without being biased towards a certain hypothesis? Earle also mentions that ECHO isn’t useful as a view of philosophy of science because “scientists can accept or reject hypotheses based on how their parts cohere and how hypotheses cohere with one another… without using a computer program” (473). With this in mind, he says that ECHO makes more sense when applied in psychology. Although I don’t agree with that, his comment regarding ECHO in the context of philosophy of science makes me question where ECHO should be applied.
Dietrich makes a good point about the distinction among facts, hypotheses, and evidence. On the surface they seem very different, but while trying to analyze their definitions I couldn’t come up with a very clear distinction among them particularly evidence and facts. If something is considered evidence is it not also considered fact? I also agree that if Thagard’s explanatory coherence should be viewed as a theory of the philosophy of science then he shouldn’t classify the definition of explanation as ‘primative.’ He should be working towards developing what it is. It seems the further we explore the explanatory coherence theory and ECHO the more words we can’t truly define (analogy, explanation, etc.). How can we apply these theories to anything if we can’t define most of these terms?
A variety of questions came up for me while reading both the responses to Thagard’s target article and his response to those critiques. In an attempt to organize my thoughts, I will begin with specific concerns and progress to more general questions.
Cohen and Sintonen (as well as Thagard in his response) address the importance of including the predictive value of hypotheses in any model of the reasoning process. As Thagard mentions, however, there may be no manner in which to quantify this principle (493). Is this a place in which probability could step in? Does there seem to be any viable way to quantify predictive value of hypotheses?
Both Giere and Thagard discuss the effect of modeling reasoning based on language, which Giere (as well as I) see as problematic. Giere states that there are “nonlinguistic representational mechanisms and judgmental strategies operative in individual cognitive agents” (476). Thagard responds that because language is central to how scientists form hypotheses and evaluate theories, it is “central to understanding the highly verbal practices of scientists” (492). I agree that language may lead us to understand their “verbal practices,” but I am not convinced that it will ever be a sufficient device for modeling human reasoning on its own. Does it seem fair to say that we are (or will ever be) conscious enough of every step of our reasoning in order to express each one verbally?
Similarly, my next small inquiry has to do with what Read and Miller mean by “accessibility” when they say that “more coherent self-systems would be those in which a given individual has more accessible self-relevant data” (486). Again, are the authors suggesting that awareness of one’s own reasoning process is necessary for “coherent” reasoning or could a person achieve this without necessarily being conscious of the links or the system? Even further, O’Rorke brings up the influence of goals and priorities in evaluation (484), and the question can then include: are we (or could we ever be) aware enough of our own motivations to even attempt quantifying them in a connectionist model?
On a larger scale, I am interested in the idea that normative rationality is even how we “should” reason. In a broad sense, if we have evolved to reason in the way that we do (which should enhance survival at least) why does it matter whether we are operating according to these “logical” (yet reductionist, in my opinion) principles? Bereiter and Scardamalia imply that we should aim to reason in order to arrive at the same “sensible results” as the computer models of normative rationality, but I am not so sure that this is as “ideal” as imagined.
On page 490, as he begins his response, Thagard writes “The best current method for psychological theorizing comes from computational modeling. From this perspective, philosophy becomes part of cognitive science; it should not seem odd to find a computer program described as part of a theory in the philosophy of science.” Is it the case that Thagard advocates for a naturalized epistemology? Collapsing philosophy of science into cognitive science seems to be a naturalizing position. It’s possible that this is obvious and I am just now picking up on it.
As for the question: do we need a “working definition” or a theory of “explanation” in order to evaluate TEC, Im not sure. Thagard’s dismissal of a philosophical analysis of explanation seems to be influenced by his belief naturalized in epistemology. He writes that, “Concepts in science and ordinary life are rarely [given as a set of necessary and sufficient conditions] (492).” Thagard does, however, give us a relatively detailed account of explanation criteria. Explanations should be quasideductive, invoke causal mechanisms, should be question oriented, and computational, to name a few. However, I do wonder if these criteria are a little unhelpful if we do not know which are necessary or sufficient for an explanation. What if an “explanation” fulfills some of these conditions but not others?
I think the peer commentary touched on many great points, several of which we also discussed in class. Throughout the peer commentary, I was most interested in a few ideas relating to ECHO’s ability to realistically mirror human reasoning and decision making:
a. First, as Bereiter and Scardamalia discuss, the role of contextual facts must be examined not only in our own decision making but also in the framework of ECHO. It seems impossible to me that ECHO can correctly accommodate contextual facts about “cultural differences, social conditions, and historical antecedents” (469). These “facts” appear to play a great deal in how humans reason, but it doesn’t seem like they can be included as easily in ECHO (this is similar to ECHO’s seeming lack of ability to include things such as emotion, bias, past experiences, etc.). If ECHO can take these contextual facts into account, how should they be weighted within the system? In addition, as Emily mentioned, what exactly are these “facts”? Are they more privileged?
b. Bereiter and Scardamalia also mention the role of contradiction. They propose that many hypotheses could exist that don’t necessarily conflict in the seemingly black and white way ECHO is programmed to simulate. What if the hypotheses explain various things with “different levels of description”?
c. Can ECHO take into account the processing limitations proposed by Chi? People cannot possibly consider the introduction of new concepts, the large quantity of evidence that already exists, and all of their interrelationships at one time. This relates to my point from last week about occurrent and non-occurrent information. It seems to be one of ECHO’s merits that it is a holistic system, but is that realistic for humans? Chi also proposes that people may come to different conclusions (and ECHO may also) depending on how they look at the evidence (in smaller subsets versus in parallel, all at once). It appears that although we may sift through information unconsciously, as Thagard proposes, we still cannot look at things in as holistic a way as ECHO can.
d. Dawes also brought up an interesting question about how the order in which evidence is presented affects our reasoning. If we find merit in this idea, how can ECHO simulate this human experience?
e. Giere: If we’re really interested in the cognitive processes that go into our conclusions, not just in the end result, or the arguments that defend these conclusions, it seems that ECHO is lacking. Giere takes issue with Thagard’s examples, proposing that “…in each case the scientist’s purpose in producing his text was not to record the thought processes by which he became convinced of the correctness of his theory, but to establish his claim on the theory and to persuade others of its value” (476). This relates not only to the goal of ECHO, but also the information that can be used in the system (the information Thagard chose to input). In the same vein, Klayman and Hogarth believe that “ECHO does not model the process of thinking, but rather its end result” (478).
f. Lastly, it appears that ECHO is a somewhat black and white system. What actions can it take to modify theories? What about synthesizing two parts of conflicting theories? (Levine)
I do think ECHO may be useful in certain contexts, but I think it will be important to define its abilities and limitations before deciding how exactly it can be applied.
Like many of the authors of the response papers, and like we discussed in our last class, I am still thinking about the context in which ECHO would be most useful and most realistic. Dietrich explains that TEC should not be considered a philosophy of science because scientists can determine accept/reject a hypothesis without using a program like ECHO. Moreover, the definitions of facts, hypotheses, and evidence are unclear and hard to distinguish from one another. These may be true, but I still personally think that ECHO/TEC would fit most appropriately as a philosophy of science just because I still have difficulty in accepting that it does mimic human psychology and how actual humans form beliefs. I remain hesitant about the lack of a clear definition of “explanation,” and after reading the response articles I am also unsure of how “analogy” fits in as an accurate principle. As I mentioned before, I think that ECHO is missing many of the other things that influence how we think – whether it is our emotional state or our attention. Dietrich states that “[Thagard] might think that understanding how science works is equivalent to understanding how human individuals work” (474) – and I think that these two are in fact very different. I think ECHO can be useful in scientific contexts, but not in terms of actual human experience. Are we willing to accept that understanding science is the same as understanding individuals, like Thagard seems to be?
I like where Reggia and Feldman are going with the notion of Bayes’ Theorem and the introduction of probability into something like TEC or the ECHO programming. From studies in Economics and with Game Theory, I am accustomed to thinking of hypotheses and potential outcomes in terms of numeric-based probabilities of each. In fact, when reading the original Thagard paper about the ECHO modeling, I took the “weights” of connections to be synonymous with their probability. I assumed that each was assigned, as opposed to strengthened by the number of hypotheses and the background programming of ECHO. Obviously, it is near impossible for one to assign a probability value to a hypothesis in TEC, and therefore I understand Thagard’s rebuttal to these critique propositions.
I am curious about the “philosophical debates about the interpretation of probability” that Thagard mentions (495). I personally find probability to be a much stronger indicator than number of hypotheses supporting a theory. Thagard says that he would be willing to use probabilities in the model if they were accurate and available, but how would one incorporate that type of math into ECHO, especially if in some cases H1 could have a true probability value but H2 is unknown?
The second half of this article presents criticisms of Thagard’s explanatory coherence model and Echo’s limitations. Bereiter and Scardamalia illustrate that explanations gain plausibility if it can intertwine with contextual facts, or facts that don’t need explaining (469). Can contextual facts be related to basic beliefs, and can Echo accommodate contextual facts? I would speculate that Echo cannot because contextual facts seem more privileged than other propositions and the propositions that created create connections should be equal in the coherentist view. Other individuals illustrate that Echo needs a way to compare contractions and competing but non-contradictory hypotheses. I thought the concept of radical restructuring was interesting because it suggests that there is a scientific mechanisms that shifts one thought to another (470). This process seems relatable to early vision in that there are mechanisms that produce sensations that can lead to certain perceptions, depending on cognitive influence. To what degree is radical restructuring a cognitively impenetrable process? Does the shift from one thought to another require a ‘set of interrelated beliefs?’ (470)
Cheng and Keane highlight my own skepticism of the term explanation that was discussed last class. They illustrate that explanation in this model comes in the form of antinational links, but because psychological processes that underlie the explanation are not clear, it is hard to tell whether they are actually explained (470). The ambiguity of explanation connects to Michelene’s question of what echo’s network of explanations represent. Michelene argues that two theorists who hold opposing views that know each others views and have the evidence for both would not come to the same evaluation of respective theories (470). This notion makes me question the presence of biases in reaching an explanation based on a hypothesis a person wants to be true. Could echo identify these biases? The fact that the representations are different suggests that even if humans have a certain orientation or frame of thought Michelene alluded too, restructuring is a cognitive process. Therefore I don’t think that echo can succeed in providing a mechanism for which conceptual changes occur be cause there may not be a single mechanism for thought.
I appreciate thagards responses because he acknowledges the limitations of echo and the theory of explanations. He does not claim that explanations are deductive, because concepts of life or science are not subject to definitions. He says instead that some explanations are quasideductive, and explanations in part function to fit schemas into situations and describe a phenomena in the context of general laws (492). This notion calls into question the degree to which echo can account for semantic relationships. Thagard doesn’t argue that echo can generate hypotheses because generation involves goals and priorities. He argues that Echo can be used to accept or reject hypothesis. However, doesn’t this process of testing hypotheses also involve cognition? Maybe the point is that they shouldn’t, as a court room shouldn’t be based on biases or jurors’ judgments. Inevitably, hypotheses of science are accepted based on empirical evidence. It becomes clear in programs such as Echo that not all propositions are created equal because evaluation involves cognition. When Thagard says,“ people do not just believe what they want although they try to find evidence for what they want to believe,” is he admitting to the fact that the process of explanatory coherence will not realistically apply or explain the process of explanation?