Playing “Fast and Loose” with Terms

One of the things I’m interested in in philosophical methodology is the “loose” use of terms. Such terms are used as argumentative shorthand. I want to explore terms that have attained metaphorical status and delineate them more precisely. I define playing “fast and loose” with a term as:

• (FAL): A term r is used imprecisely when the argumentative weight r must carry within the argument is disproportionate to the actual weight presented for the term either within the argument itself or within one’s peer community.

(FAL) might cover using terms out-of-context and misusing terms against their actual (or disputed) use within one’s peer community.

Objection: Isn’t this is what analysis is—the percise use of concepts and language? Why not just say people are doing bad analysis when they use terms loosely?

Answer: People are doing poor analysis, but (FAL) helps focus on the poor use of terms within the broader analytical apparatus.

An example of (FAL) in action is a recent blog post by Kvanvig. In his post Kvanvig used the term “reflective equilibrium” loosely. I raised this to his attention in reply #7 found here. He clarified his use of reflective equilibrium as not a method of justification but as shorthand for the result of achieving consistency between general and particular judgements. However, reflective equilibrium is a method of (moral) justification. It is a further misappropriation to use it in the epistemic sense; or, at a minimum, it is a poor choice to use it epistemically because it “takes on” an insurmountable number of deficiencies (i.e. problems within no known solution in belief revision, objections to coherentist methodology and problems with intuitions). Reflective equilibrium is not the result of consistency between general and particular judgments. This blurs the distinction between judgments and principles and wide and narrow versions of the method. There are more problems with his use of the term, but my question is this: Why didn’t he just say he was talking about prodding his students toward a search for consistent beliefs?

Answer: Because Kvanvig was using the term to carry weight in his argument that was disproportionate to its actual weight or value. 

(FAL) is a tempting technique because it’s an easy way to add force to your argument without much work. When this occurs over time, such terms become the bloated technology stocks of the late 1990’s. And, like the stock market, one’s peer community can buy into the hype and perpetuate the false perception. There are probably a host of justifications for the habitual use of (FAL) terms: why raise such “minor” issues, everyone seems OK with using it loosely, and so on.

Ironically, (FAL) must itself be revised because a term within the definition is another term used loosely by the analytic community (i.e. weight). When I get back from the upcoming workshop I plan on exploring “weight”. It is used in many different contexts across a spectrum of sub-disciplines within philosophy. What is going on when people use the term “weight” as something that carries argumentative force? What does it mean to say that X outweighs Y? What does the use of this term presuppose? Does the use of the term vary across philosophical concepts (e.g. intuitions, beliefs, desires, reasons, judgments, principles, and so on)? Those are a few of the questions that might be worth exploring.

Bonevac’s “Reflection Without Equilibrium”

In “Reflection without Equilibrium” Dan Bonevac argues against reflective equilibrium (RE) as outlined by Rawls. Dan concludes RE is a procedure that never terminates in a finite amount of time. As a result, RE must be revised in a pragmatic, intuitionist direction. I will argue this conclusion is based on a misunderstanding of Rawls’ constructivist model and how this model solves what’s known as the priority problem—assigning weight to a plurality of competing values.

First, let’s setup one of Dan’s arguments. In Theory of Justice (p.45) Rawls contrasts intuitionist versus constructivist solutions to the priority problem. Dan takes Rawls’ comments to imply that people can choose to make the moral facts as simple or as complex as they want. If this were true, according to Dan, wanting the priority problem to have a solution would make it have a solution. This would be a strike against the intuitionist because he believes that the plurality of competing values cannot be simplified; there are no higher order principles that can be ranked and consulted to determine the outcome and settle the issue between competing values or moral facts. Dan takes the quote from Theory (p. 45)  to imply that we could choose to make the moral facts simple, thereby defeating the intuitionist’s complexity thesis. From this line of reasoning the following argument is laid out on page 19:

1. Choices in the original position determine the principles of justice.
2. The principles of justice chosen in the original position determine the moral facts.
3. People in the original position would choose to make the moral facts simple.
4. So, moral facts are simple.

To show how deficient such reasoning, Dan runs a parallel argument about a batter wanting to hit a home run: (i) if a batter’s choices determine the trajectory of the bat and (ii) the trajectory of the bat determines if the ball goes over the wall, then (iii) every batter would chose to hit a home run on every pitch, so (iv) every batter would hit a home run on every pitch. Dan analyzes premises (1 and i) and shows how many factors determine whether the choices a batter makes determine the trajectory of the bat or whether choices in the original position determine the principles of justice. He weakens this premise to account for these factors (i.e. choices, under ideal conditions, contribute to determining…).

Next, Dan skips premise (2) and argues against premise (3). This is a mistake because (2), while stated explicitly by Rawls, is take out of context and misused by Dan. (2) is not a premise Rawls uses in an antirealist argument against the independent existence of moral facts. Rawls is not constructivist in terms of being an antirealist; instead, Rawls is a constructivist along Kantian lines. He believes certain criteria can be used by rational people in deliberation to reach agreement or disagreement about what is the case. Rawls thought people in the original position could reach agreement in judgments. The principles chosen in the original position and their coherence with moral judgments produces an outcome that reflects our moral sensibility.

Rawls was not arguing that people can fashion the moral facts in any way they see fit. The moral facts are not constructed in the sense of being infinitely pliable (thus, they are able to be made simple or complex based on how we want them to turn out).  Instead, the moral facts are constructed because simple facts, the existence of which Rawls or Kant would not deny, are determined to be moral facts by principles that assign them weight as reasons. This assigning of weight solves the priority problem. The moral facts are then used, within RE or the categorical imperative procedure, to support certain conclusions. Moral facts are not infinitely pliable nor are they fixed ethical truths. They are simple facts principles have selected as relevant from a moral point of view. I would reformulate the argument as follows:

1. Choices in the original position determine the principles of justice.
2. The principles of justice chosen in the original position determine the moral facts.
3. People in the original position can agree (or disagree) about the moral facts.
4. So, moral facts are objective.

Another consequence of Dan misconstruing the form of constructivism proposed by Rawls is that his argument against RE reaching equilibrium is on less sure footing. That is, it is wrong to associate the set of stable judgments with ethical truths. The target of reasoning for Rawls is not ethical truths but objectivity in the sense of agreement that mirrors our sense of justice.

Moral Heuristics

In my last post I claimed intuitions are often used (or should be used) like heuristic devices. As shortcuts of cognition, intuitions function as intellectual seemings that quickly move the agent from perception to judgment. This leap occurs without explicit analysis or sifting through evidence; instead, the agent references a rule of thumb. A rule of thumb is a generalization about what to do, think or feel in a certain situation. These generalizations are often highly intuitive and have the attractiveness of being common sense. The problem is that intuitions generated by and used like heuristics often reflect errors and biases.

My research is primarily in the area of reflective equilibrium. This method of moral justification is often charged with being intuition-laden, so the growing literature on experimental philosophy and intuition is of sincere interest. In the journal Behavior and Brain Sciences I came across an article about moral heuristics. It includes a primary article by Cass R. Sunstein and an extensive peer commentary on his article. Sunstein discusses heuristics and reflective equilibrium. Commentary related to this topic include: Peter Singer’s “Intuitions, Heuristics, and Utilitarianism,”  Edward Stein’s “Wide Reflective Equilibrium as an Answer to an Objection to Moral Heuristics,” and Philip Tetlock’s “Gauging the Heuristic Value of Heuristics.” In discussing the link between heuristics and morality Sunstein states:

Much of everyday morality consists of simple, highly intuitive rules that generally make sense, but that fail in certain cases. It is wrong to lie or steal, but if a lie or a theft would save a human life, lying or stealing is probably obligatory. Not all promises should be kept. It is wrong to try to get out of a longstanding professional commitment at the last minute, but if your child is in the hospital, you may be morally required to do exactly that (p. 531).

I plan on posting on this article in the future, but for now I wanted to bring it to your attention. The article can be found at the following site within Cambridge journals.

Challenging How Intuitions are Used

One required reading for Ernest Sosa’s presentation is Jonathan Weinberg’s “How to Challenge Intuitions Empirically Without Risking Skepticism.” In this paper, Weinberg argues intuitions are hopeless as sources of evidence. A source of evidence is hopeful to the extent it can detect and correct errors in its use. Vision is used as an example of such a capacity. Though vision is fallible it has checks and balances that mitigate against its fallibility (i.e. provide data on when it is fallible and compensate for its fallibility). So, perception is a “hopeful” source of evidence. Weinberg suggests that if a source of evidence is “hopeless” it should not be trusted.

(H): Any putative source of evidence that is hopeless ought not be trusted (p. 327).

Why should we affirm (H)? Without philosophical methods, and the evidence they are based on, possessing self-correcting abilities errors remain undetected and an entire discipline can get off track, chasing futile lines of reasoning. There is no good way to resolve disagreements between theorists when they are using hopeless sources of evidence. Positively formulated, there are four ways our practices can avoid being hopeless. These criteria are considered sources of hope.

• External Corroboration: Verifiable against fact and theory.
• Internal Coherence: Agreement both within and across subjects. Ability to filter outliers, inconsistencies and contradictions.
• Detectability of Margins: Zones of reliability are clearly demarcated.
• Theoretical Illumination: Understanding why the thing works when it works and why it doesn’t work when it doesn’t.

Weinberg argues intuition fails to affirm principle (H) for reasons I won’t rehearse in detail. Basically, intuition fails to consist in the aforementioned sources of hope. I will argue that intuition taken as a stand-alone source of evidence is hopeless, but intuition is hopeful when embed within the practice of reflective equilibrium.

So that I’m not talking past Weinberg (a complaint he levels against defenders of intuition), it’s important to keep the target before us. Weinberg is targeting a narrow range of the use of intuitions (e.g. Gettier cases, Searle’s Chinese room, Davidson’s Swampman). Weinberg is not an outright skeptic of intuitions, hence the title of his paper, rather he is skeptical of the way philosophers have used intuitions. He claims philosophers use intuitions as foundational bits of evidence, which do not require further defense or empirical justification. Specific intuitions about specific cases are used to argue for or against a particular claim. If philosophers actually use intuitions in this way, then I agree with Weinberg that the practice is hopeless. I also agree with Weinberg that, “hope does not always come from intrinsic aspects of the source of evidence itself so much as from the particular practices of using it” (p. 331). Yet, Weinberg fails to argue that Searle and Davidson were using intuition in the limited sense, as foundational bits of evidence, to support their philosophical theories. Without this argument another plausible interpretation of philosophers’ use of intuition is that intuition is being used pragmatically.

Philosophers may be using intuitions in a way that roughly mirrors reflective equilibrium (RE). This becomes clearer as one zooms in and out of the arc of a philosopher’s argument, as one views her whole body of work and as one traces how her thought progressed based on critical dialogue with other philosophers. Disagreement prompts revision and refinement. In such a practice, intuitions serve as revisable bits of data. Confidence in particular intuitions about particular cases may vary depending on how an intuition relates to the whole set of data. Intuitions feed the inference machine by providing a practical starting point for the formation of a robust theory or form of epistemic justification. I could make a stronger case for this point by specifically showing how Searle and Davidson are using intuitions in this way, but, for now, the mere plausibility that this is what is happening is enough to cast doubt that both proponents and opponents of intuition know what they are doing when they are using intuitions. I hold that they are using intuitions as useful heuristics or first approximations. These approximations often alert philosophers and their audience that something else must be accounted for—that there is a need for additional reasoning, theorizing and analysis.

Finally, I will argue intuitions are hopeful when used as bits of evidence within the method of reflective equilibrium. Intuitions gain their hope within RE because RE is a practice that exhibits all four criteria of hope.

• External Corroboration: Empirical facts and a wide range of theories are considered; adjustments to one’s total set of evidence are made in light of these considerations.
• Internal Coherence: To quote Rawls, the goal of reflection is, “a matter of the mutual support of many considerations, of everything fitting together into one coherent view” (TJ, p. 21).
• Detectability of Margins: Conditions prone to error are filtered out as intuitions become considered judgments. Initial confidence in what seems the case is refined as intuitions are thrown out that were made under conditions of emotional duress, conflicts of interest and other distorting factors. Further, the margins of correctness are determined as inference proceeds and inconsistencies between judgments, principles and theories are spotted and corrected.
• Theoretical Illumination: Logical inference, belief revision and folk psychology provide a theoretical canvas to explain the success and failure of trying to reach a mutually coherent set of judgments, principles and theories.

My contention is that philosophers implicitly, and poorly, use a method that approximates reflective equilibrium when deliberating and justifying their claims. Instead of going back and forth about whether armchair philosophy is valid or whether what seems to be the case varies according to socio-economic demographics our efforts will be better spent refining what philosophers naturally do. That is, the method of reflective equilibrium needs to be made more rigorous or it needs to be abandoned. But, if it is abandoned, a suitable alternative needs to be put in place. Currently, I am not aware of a more robust method of justification. Such a method is not found by arguing about intersubjective agreement, trying to locate the source of origination of intuitions or trying to determine when intuitions work as stand-alone (foundational) sources of evidence. Rather, our efforts will be better spent refining the method of reflective equilibrium.

Accepted to Philosophical Methodology Workshop

I just got back from my honeymoon in Kauai (what an amazing island), so this is a delayed announcement. I’m pleased to announce that I was accepted to attend a workshop on philosophical methodology. I’m deeply grateful for this amazing opportunity to work with professors and grad students from across the country. It will be an intensive five days of seminars hosted by professors. Information about the workshop is found here. In the next month, I’ll post comments on assigned readings for some of the seminars.

New Papers Posted

I’ve posted two new papers on my Website. Click here to go to the papers. The first paper is the result of a directed reading course with William H. Shaw. The second paper is the result of studying David Davies’ book Art as Practice. Both papers handle issues around the method of reflective equilibrium. In the first paper I develop an alternative account of reflective equilibrium, one designed to address defects in standard interpretations of the method. The second paper argues that David Davies’ built his performance theory in aesthetics on a faulty foundation. To shore up the methodological principle supporting his theory I make important revisions to his principle.

Reflective Equilibrium and Machine Ethics

Anderson and Anderson have designed a decision procedure to make a robot act ethically. This project is detailed in a paper found here. In this post, I will claim that Anderson and Anderson’s misappropriation of Rawls’s reflective equilibrium brings out an important objection to their project.

First, I will provide a rough outline of their project. For Anderson and Anderson the best way to implement ethics into a machine involves ethical principles. In the quest for a principle that could guide a robot to act ethically in cases where prima facie duties conflict, the system is presented with test cases. These cases are used as training data. They provide answers about ethical dilemmas where ethicists are in agreement about the right course of action. Based on these cases the machine generates a principle thought to cover the rest of the possible cases (in a situation comprised of three duties and two choices). If the machine provides a correct answer for the remaining cases, measured against agreement among ethicists about the cases, then the decision principle is complete and consistent. After stating the principle the machine discovered, Anderson and Anderson concluded:

This philosophically interesting result lends credence to Rawls’ “reflective equilibrium” approach—the system has, through abstracting and refining a principle from intuitions about particular cases, discovered a plausible principle that tells us which action is correct when specific duties pull in different directions in a particular type of ethical dilemma (p. 2).

Have Anderson and Anderson really discovered a result that “lends credence” to reflective equilibrium, or have they, by quasi-appropriating reflective equilibrium, attracted attention to a flaw in their approach? 

One of the objections to reflective equilibrium is that it is a rigged procedure (see Brandt, 1979). That is, lending initial credibility to intuitions about specific cases is thought to taint the outcome; credence placed in the input data, the considered judgments about specific cases, determines the outcome of the procedure. This problem finds acute expression in Anderson and Anderson’s absolute confidence in the agreement of ethicists about particular dilemmas. The cases are fixed starting points and final arbiters of the success of the procedure. To escape the starting-point dependence objection to reflective equilibrium philosophers emphasize that judgments about particular cases are vulnerable to revision throughout the procedure (see Scanlon, 2003). Belief revision involves “working back and forth” between the judgments and the principles to reach agreement between them, sometimes revising judgments and sometimes revising principles. Because Anderson and Anderson used fixed intuitions, and sought to accommodate the intuitions by bootstrapping up to a principle, the decision procedure they employed does not even count as an example of narrow reflective equilibrium–when judgments and principles are altered to achieve the best ’fit’ between them. Anderson and Anderson have used a one-way arrow of revision, holding immune to revision the judgments of experts. Thereby, they rightly observed about the output of the procedure, “clearly, this rule is implicit in the judgments of the consensus of ethicists” (p. 2). The result is that the machine has not engaged in rational deliberation about moral situations, it has merely codified what was already implicit in the intuitions about the cases. Thus, the procedure begged the question against the outcome. Does such a method represent sound ethical deliberation? Or, could a method other than principle-based reasoning about duties provide a better alternative? 

I explored using utilitarianism as an impetus for designing an ethical robot in a paper found here. I have also considered using utilitarianism to inform the design of better, more ethical, decision-theoretic machines in the draft of a paper found here. Yet, I am not convinced either of these approaches are the best way forward for machine ethics. Anderson and Anderson are onto something in wanting to appropriate reflective equilibrium in designing ethical agents. Perhaps, the most robust approach to implementing ethics in a machine would involve a faithful appropriation of reflective equilibrium—one that achieved narrow equilibrium and progressed toward wide equilibrium. Wide reflective equilibrium involves consulting background theories thought to shed light on the strengths and weaknesses of the principles and judgments in one’s belief set (see Daniels, 1996). As Rawls has characterized the procedure, “one seeks the conception, or plurality of conceptions, that would survive the rational consideration of all feasible conceptions and all reasonable arguments for them.” (Collected Papers, p. 289). This may seem an ambitious approach to machine ethics, but what other methodology could be more justified?

Good Philosophy Presentations

I just returned from the Society for Student Philosophers (SSP) annual conference at UT, Austin. It was a well-run event with good presentations and discussions. I wrote a comment elsewhere about how good presentations involve multi-media elements. At one level, the goal of a good presentation is the same as the goal of a good argument.

(G1) A sound argument is an an argument that is valid and contains only true premises.

Added to G1 is G2 which states:

(G2) An argument is valid if and only if it is necessary that if all the premises are true, then the conclusion is true.

Soundness and validity are basic standards of what makes an argument good. A good philosophy presentation should be held to the same standard. The presenter should state a conclusion, bolster the conclusion with premises that necessarily lead to the conclusion, and make evident the truth of the premises. A further criterion of a good argument is cogency. A.P. Martinich in Philosophical Writing states cogency as:

(G3) A cogent argument is a sound argument that is recognized to be such in virtue of the presentation of its structure and content.

The cogency requirement is a requirement of recognizability: the soundness (and validity) of the argument should be easily recognizable. If adequate evidence for premises is not presented, then even a sound argument can miss the mark of cogency. Evidence for a premise should make it obvious that the premise is true. The structure of the argument must also be apparent. The relations between premises and their tie to the conclusion should be well-articulated. Cogency is a place where many philosophy presentations miss the mark. With this in mind, there are two elements of a good presentation that must be held in tension:

(A1) A philosophy presentation is accurate if it contains a sound argument conveyed with precision.

Accuracy is held in tension with cogency:

(A2) A philosophy presentation is good if and only if it engages an audience with the recognizability of its accuracy.

Both A1 and A2 involve performative elements. All aspects of a presentation should be filtered through A1 and A2. The basic question is, “Does this medium aid in reaching the end of a good argument presented well?” The overall goal is to achieve the right balance between accuracy and cogency. I’ve been wrestling with how to achieve the right balance between A1 and A2 since the conference.

Most students at the conference erred on the side of accuracy to the determent of cogency. Most students read a shorted version of their extended paper. This facilitated achievement of A1, but it left the audience groping for A2. Students who read their paper tended to overwhelm the audience with a complex argument that didn’t make apparent how the evidence presented was supporting the premises. Put simply, too much information was thrown at the audience. This meant that ‘question time’ often involved clarifying questions instead of sharp questions addressing the support (or lack thereof) for the premises.

Overall, I was pleased with my presentation. I was glad I put more time into the slides and allowed that process to inform a tightening of my argument. It also served the end of peaking Scanlon’s interest and led to an interesting debate with him in the break-room over the details of Reflective Equilibrium. When one talks extemporaneously from bullet points (using PowerPoint or an outline in the form of a handout) it can facilitate achieving A2, but it can often do so at the expense of A1. I wish I would have achieved more accuracy by having the premises of my argument on a one-page handout. This would have also achieved a greater recognition of the soundness of my argument (or lack thereof).

Scanlon’s presentation was amazing. He talked extemporaneously off bullet points on an overhead projector. He did this because he’s Scanlon and he has been thinking and presenting for more years than I’ve been alive. But, I think he served as an excellent example of talking simply about a complex subject (rationality and reasons). His talk was engaging because he conveyed his ideas with both accuracy and cogency. The media he used did not get in the way of the ideas he was conveying. However, he could have moved his bullet-points to PowerPoint slides and his presentation would have been more engaging.

All that to say, I’m still wrestling with the best way to hold A1 and A2 in tension, how to achieve the right balance between these two vital elements in a good presentation.