Thursday, June 19, 2014

A Cynical Argument for the Liberal Arts, Part Fifteen

Previously: Part Zero ... Part Fourteen

After being derailed last time by a quotation, I'll take up the question of constructive deconstruction: how can the corrosive truth-destroying effect of Cynical "debasing the coin of the realm" lead to improvements within a system? Note that a system (loosely defined: society, a company, a government) is the required  'realm' in which to have a 'coin.' My modern interpretation is official or conventional signalling within a group. Waving hello and smiling are social coins of the realm. Within organizations, a common type of signal is a measure of goal attainment, like quarterly sales numbers. It's this latter, more formal type of signal, that is our focus today.

A formalization of organizational decision-making might look like this (from a slide at my AIR Forum talk):


Along the top, we observe what’s going on in the world and encode it from reality-stuff into language. That’s the R → L arrow. This creates a description like “year over year, enrollment is up 5%,” which captures something we care about via a huge reduction in data. R → L is data compression with intent.

As we get smarter, we build up models of how the world behaves, and we reason inside our language domain about what future observations might look like with or without our intervention. We learn how to recruit students better under certain conditions. When we lay plans and then act on them, it is a translation from language back into the stuff of reality—we DO something. By acting, we contribute to the causes that influence the world. We don’t directly control the unfolding of time (red arrow), and all pertinent influences on enrollment are taken into account by the magical computer we call reality. Maybe our plans come to fruition, maybe not.

Underlying this diagram are the motivations for taking actions. Our intelligence, whether as an individuals or as an organization, is at the service of what we want. It's an interesting question as to whether one can even define motivation without a modicum of intelligence--just enough to transform the observable universe (including internal states) into degrees of satisfaction with the world. As animals, we depend on nerve signals to know when we are hungry or in pain. These are purely informational in form, since the signals can be interrupted. That is, there is no metaphysical "hunger" that is a fundamental property of the universe--it's simply an informational state that we observe and encode in a particular way. In a sense, it's arbitrary.

For organizations, analogs to hunger include many varieties of signals that are presumed to affect the health of the system. Financial and operational figures, for example. These are often bureaucratized, resulting in "best salesman of the quarter" awards and so forth. An essential part of the bureaucracy is the GOAL, which is an agreed-upon level of motivation achievement, measured by well-defined signals. An example from higher education is "Our goal for first year student enrollment is an increase of 6% over three years."

A new paper from the Harvard Business School "Goals gone wild," by Lisa D. Ordóñez, Maurice E. Schweitzer, Adam D. Galinsky, and Max H. Bazerman, provocatively challenges the notion that formal goals are always good for an organization. This itself is a cynical act (challenging established research by writing about it), but the paper itself is a great source of examples of how Cynical employees can react to a goal bureaucracy in two ways:

  • Publicly or privately subverting goals through actions that lead to real improvements in the organization, or 
  • Privately debasing the motivational signals that define goals for personal gain.
The first case is desirable when the goals of an organization, when taken too seriously, are harmful to it. Too much emphasis on short-term gains at the expense of long-term gains is an example. This positive reaction to bad goals is not the topic of the paper, but proceeds from our discussion here. The second case is ubiquitous and unremarkable: any form of cheating-like behavior that inflates one's nominal rank, reaping goals-rewards while not really helping the organization (or actually harming it).


Earlier, I referenced a AAC&U survey of employers that claimed they wanted more "critical thinkers." Employees who find clever ways to inflate their numbers is probably not what they have in mind. Before the Internet came around, I used to read a lot of programming magazines, including C Journal and Dr. Dobbs. One of those had a (possibly apocryphal) story about a team manager that decided to set high bug-fixing goals for the programmers. The more bugs they found and fixed, the higher they were ranked. Of course, being programmers, they were perfectly placed to create bugs too, and the new goals created an incentive for them to do just that: create a mistake, "find" it, fix it, get credit, repeat. This is an example of organizational "wire heading." 

From the Harvard paper's executive summary, we can see problems that an ethical Cynic can fix;
The use of goal setting can degrade employee performance, shift focus away from important but non-specified goals, harm interpersonal relationships, corrode organizational culture, and motivate risky and unethical behaviors.
Armed with the liberal-artsy knowledge of signals and their interception, use, and abuse, AND prepared with an ethical foundation that despises cheating, an employee has some immunity to the maladies described above. Of course, this depends on his or her position within the organization, but generally, Cynical sophistication allows the employee to see goals as what they really are--conventions that poorly approximate reality. This is the "big-picture" perspective CEO-types are always going on about. Moreover, an ethical Cynic who is in a position of power is less likely to misuse formal goal-setting as a naive management tool. 

The paper itself is engagingly written, and is worth reading in its entirety. With the introduction above, I think you'll see the potential positive (and negative) applications of Cynical acts. Informally, the advice is "don't take these signals and goals too seriously," which is a point the original Cynics repeatedly and dramatically made. More formally, we can think of a narrow focus on a small set of goals as a reduction in computational complexity, in which our intelligent decision making has to make do with a drastic simplification of the world. Sometimes that doesn't work very well, and the Cynics are the ones throwing the plucked chickens to prove it.

Next: Part Sixteen

Friday, June 06, 2014

A Cynical Argument for the Liberal Arts, Part Fourteen

Previously: Part Zero ... Part Thirteen

Because Cynical "debasing the coin of the realm" corrupts the way in which individuals or organizations are aware of the world, it may be hard to see how one gets beyond this destruction. Now I would like to speculate on how we can think of Cynicism as a constructive method.

Cynical attacks on categorical ways of knowing (I called them signals earlier) challenge the categories by turning something into its negation. A counterfeit is, and is not, a coin. When I started thinking about using a destructive method of constructing, the Sherlock Holmes quote came to mind. This led me down a rabbit hole that ironically illustrates the point. When I searched for the quote, I found the following attributed to Sir Arthur Conan Doyle:
Once you eliminate the impossible, whatever remains, no matter how improbable, must be the truth.
I wanted to be sure, and it took some time to track down a reference. One site put these words in the mouth of Doyle's Sherlock Holmes in "The Adventure of the Beryl Coronet". However, this doesn't seem to be the case. After more work I found the complete works of Doyle in a single text file here. The word 'eliminate' is only used twice in the whole body of work. Here's the one we want, in the first chapter of The Sign of Four,  with Watson speaking first:
"How, then, did you deduce the telegram?"
"Why, of course I knew that you had not written a letter, since I sat opposite to you all morning. I see also in your open desk there that you have a sheet of stamps and a thick bundle of postcards.  What could you go into the post-office for, then, but to send a wire? Eliminate all other factors, and the one which remains must be the truth."
[Update: I subsequently found the actual quote by searching for 'improbable'. Leaving the rest of the post unchanged.]

Now it's possible that Doyle used the first quote outside of a novel, but the story becomes odder when we look at the record. Using Google's ngram server, I searched for "eliminate the impossible" and "eliminate all other factors." Here's the result:

Red line is "eliminate all other factors", blue is "eliminate the impossible"
Tantalizingly, this has the common quote coinciding with the publication date of The Sign of Four, according to this site. The oldest edition I could find was a book (not the original magazine, in other words) scanned from University of Michigan's library:


Its publication date is "MDCCCCI," or 1901. I tried adding the "you" in front of the first quote, and got no hits on the ngram server, but it may have a limit on how long phrases can be. After more searching I found a scanned copy of the original 1890 Lippincott's Monthly Magazine. It has the same text as the book shown above. This would seem to rule out a change in the language between the original magazine publication in 1890 and subsequent compilation in 1901.

Google Books results show that there are 19th century instances of "eliminate the impossible," which probably explain the eruption of the blue line in the graph before the (book) publication of The Sign of Four in 1901. I looked for other partial phrases too, like "no matter how improbable." These didn't turn up Doyle, but other treasures. From Doing Business in Paris:


 and this one:
The sentiments of both of these is that untruths can propagate wildly, given the right conditions, which may be what we're seeing here, and a good reason for Cynical weeding out.

Moving forward in time, I scanned 1900-1920 with Google Books, and found a magazine called Woman Citizen with this text from April 26, 1919.

The offhand rephrasing switches "Eliminate all other factors" with "Eliminate the impossible." Six years later, we find:

This has the "however improbable" drama that Doyle didn't put in the mouth of Holmes. The attributed quote finally appears in full form in 1949:

Not only is the hyperbolic "however improbable" in appearance, it's in italics for extra effect. The book seems to be an edited volume by various authors of Doyle/Holmes-related biography, anecdotes, and lit crit, although all I have to go by is the random samples that the stingy owner of copyright allows Google to produce. Worldcat coughed up only five copies, none of which are online. Let us allow the matter to come to rest here.

By eliminating possibilities, we come to the tentative conclusion that Doyle's famous quote is actually due to fan fiction, which can easily be seen as a Cynical debasement of the original. This example illustrates this by switching out Doyle's concise "Eliminate all other factors, and the one which remains must be the truth" with the nearly breathless "Eliminate the impossible, and whatever remains, however improbable, must be the truth." The fact that the latter one is much more well known, and attributed to Doyle is a clear debasement of the latter's words, and is a minor disturbance in the reality of anyone who believes the wrong version. The famous quote is, and is not, by Sherlock Holmes.

It takes a lot of work to separate signal from noise, which makes Cynics dangerous. Now what about that construction? That will have to wait until after the rabbit soup.

Next: Part Fifteen

Thursday, June 05, 2014

A Cynical Argument for the Liberal Arts, Part Thirteen

Previously: Part Zero ... Part Twelve

Here we continue the discussion of Cynicism in higher education, where I continue to use the charge "debase the coin of the realm" as the tool for analysis. In a previous installment I said that any statement of fact was a violent act. Allow me to pick up that idea here, since much of college consists of listening to statements of fact.

We constantly observe the world through our senses of internal and external affairs, and some of these we encode into language, as in "the cat just knocked over the vase." This entails data compression, since we don't have time to describe everything about the cat and the vase or the irrelevant features of the situation. But it's such a convenience that we may forget that it's just language, just a crude approximation of what was actually observed in that moment. If you'll allow me a neologism, I'd like to refer to compressed data as 'da'. So we pass da back and forth, decompress it internally, ignore the inevitable errors, and this becomes a high value "coin of the realm." If you tell me "the bridge is washed out ahead, but if you turn at the gas station you can get to town that way," it's useful da. However, passing falsehoods along debases this coinage, and lying is something we tell kids not to do.

A statement like "seasons are due to the fact that the Earth's axis of rotation is tilted relative to its orbital plane" require work to decompress. Much work in college is just to build up an ontology that lets a student associate new ideas in useful ways, sometimes by translating them into pictures. A movie that illustrates the principle in the sentence will be more effective than that sentence because moving pictures can directly simulate motion of the Earth and sun, and a viewer can create his or her own da from it. I assume that much of the time students don't know what we professors are talking about. It's just so much da-da, and that represents a failure to teach or learn or both.

Statements of fact that are inaccessible for technical reasons may not be threatening simply because it's a foreign language, and easily dismissed. On the other hand, much of what students learn in the humanities will challenge the way they think about the world. Religious fundamentalism running up against modern biology is an obvious example. But the critical theory that accompanies much of the humanities is more insidious, and can undermine the whole project of reasoning and knowing with a challenge from relativism. A student who internalizes this may not know what to believe by the end of it. There is something good in this idea that there is not always a single correct answer to a question, and that the process of reasoning is more valuable than any one product of it. It seems that the opposite is generally trained into students in the typical K-12 curriculum, where advancement may depend on a standardized test.

To explicate the process of deconstructing truth, allow me to re-use the example of Diogenes tossing the plucked chicken at the feet of Socrates, after the latter declared man to be a featherless biped. How does this trick work?

The Cynical method of attack in this case comprises an action that demonstrates a contradiction between the real world and the way we are told to perceive it. A modern example is found in "The Serial Killer Has Second Thoughts: The Confessions of Thomas Quick," which I will take at face value for the purposes of this argument (feel free to debase that value). A mentally confused man confesses to brutal crimes, and becomes a celebrated serial killer. Then it turns out that he couldn't have committed the crimes he confessed to. Many people have been convicted of crimes that they didn't commit. But it seems that he willingly participated in this fiction, which makes him a fowl-flinging Cynic. The coin of the realm debased here is the operation of the justice system and media (in Sweden in this case) and its accurate representation of reality. Like most public Cynical acts, it comes at a high price.

In order to construct such an epistemological attack on a system, one needs to find categories that are incompatible and by acting produce examples that occupy both categories. For example, the conception of worker-as-machine, which is an employer's point of view, conflicts with worker-as-human, which is the employees' point of view. A worker strike is an incompatible categorization: workers who are not working, and therefore a Cynical chicken toss. Once you see the trick, other examples become obvious: art that is not art (Duchamps), slave-as-beast versus slave-as-human, animal-as-meat versus animal-as-living-creature, Earth-as-resource versus Earth-as-home.

Some of the major discoveries of the world are paradoxical category-defiers. The discoverers of the calculus used infinitesimals, which are quantities 'infinitely small but not zero.'  The only thing infinitely small can mean is zero, so this is contradictory, and yet it is exactly the idea needed to unlock differential calculus. Or the Copenhagen interpretation, in which reality acts like probability waves collapsing into real particles. Or Einstein's category-breaking conceptions of space and time (how can time operate at different rates, when time determines what a rate is?). Or the Dirac Delta, which is essentially a box that has zero width, is infinitely tall, and contains one unit of area! The Liar Paradox "this statement is false" in the hands of Russel and Turing and Gödel caused a rethinking of the fundamentals of mathematics and set logical limits on what we can know. These we should probably consign to 'cynical' rather than 'Cynical' status, since they are arguments rather than physical acts, but this may be splitting hairs.

More common than public abuse of our feathered friends is the secret use of contradicting categories for personal gain. When these become public, they may serve as inadvertent Cynical examples. For example, Slate's "Here’s the Awful 146-Word “Essay” That Earned an A- for a UNC Jock". Quote:
The University of North Carolina–Chapel Hill has already been embroiled in a scandal for allowing its athletes to enroll in fake courses for easy credit. 
A university's motivation for recruiting a top athlete is different from its motivation for recruiting a top student, yet NCAA rules require that athletes also be successful students (there is no requirement the other way around). When forced to coincide for purposes of classification, the overlap "student-athletes" seems to often include examples that fit the latter but not the former category. Actually producing these examples is a Cynical act, even though it's not intended to be public (quite the opposite). There is a straightforward way to debase the coin of the realm (the meaning of college grades). Unfortunately for these Cynics, sometimes people find out make it public.

It seems that private Cynicism for gain, as opposed to public Cynicism, needs a name. I will refer to it as crypto-Cynicism, meaning hidden. The name seems appropriate, since this is a lot of what spies do, like assuming the identity of someone else, forging documents, getting people to trust them who shouldn't, or hiding a real message within an apparent one.

The examples show that Cynicism (as I have interpreted it) remains a powerful force for change, and that it doesn't come with ethical instructions. Colleges that teach the students the deepest forms of subversion, like deconstruction, relativism, critical theory, and so on (these overlap) hand over to the young minds solvents that can turn the world--and their own minds--to goo. Since becoming secular, colleges shy away from construction part, which is unfortunate. We could be more intentional about the existentialist project that should follow the deconstruction--finding a personal meaning to construct out of the goo. But aside from that, there is a more troubling question.

Perhaps what employers want is really two things: (1) a class of graduates who are smart enough to pick up new tasks of varying complexity and fill the technical and social demands of being a machine-part in an organization, and (2) a smaller cadre of crypto-Cynics who are willing to break rules in order to get ahead. If so, then the unsettling conclusion is that a two-class system is exactly what's needed. The first is trained to follow rules and keep their heads down, mastering jobs that real machines will eventually take. The second provides motivation and vision uncluttered by the norms of society, its laws, science, or even a common understanding of the world, to serve as leadership. The spectrum of this second group would range from normal humans to psychopaths and mystics, who succeed or fail depending on their environment. The contrast between these two types (1) and (2) can perhaps be seen in stories like this one from the Huffington Post: "For-Profit College Enrolls, 'Exploits' Student Who Reads at Third-grade Level" in these two quotes:
A librarian at a southern California campus of Everest College abruptly resigned last week, deeply upset that the for-profit school had admitted into its criminal justice program a 37-year-old man who appears to read at a third-grade level.  
versus
Everest is owned by for-profit giant Corinthian Colleges, which is facing a lawsuit for fraud by the attorney general of California and is under investigation by 17 other state attorneys general and four federal agencies. [A senior administrator] at Corinthian, told me today that the campus believed it was appropriate to take a chance on admitting the student.
There's a natural fix to this particular problem: require that the leadership of the company must come from its own graduates.

[Part Fourteen]