Wednesday, February 14, 2018

Philosophy Relies on Those Double Majors

Happy Valentine's Day! Because I love you, here are some statistics.

Following a suggestion by Eric Winsberg, I decided to look at the IPEDS database from the National Center for Education Statistics to see how commonly Philosophy is chosen as a second major, compared to as a first major, among Bachelor's degree recipients in the United States.

I examined data from all U.S. institutions in the database, over the three most recent available academic years (2013-2014, 2014-2015, and 2015-2016), sorting completed majors by the IPEDS top-level major categories (2010 CIP two-digit classification), except replacing 38 "Philosophy and Religious Studies" with 38.01 "Philosophy". Separately I downloaded the grand totals of completed first and second majors across the U.S.

In all, across all institutions, 5,680,665 students completed a first major. Of these, 289,639 (4.9%) also completed a second major. Not many students earn second majors!

However, the ratios vary greatly by discipline. In all, 24,542 students earned a Philosophy major, of which 5,015 (20.4%) earned it as a second major. At a minimum, then, 20% of Philosophy majors are double majors. If half of those double majors choose to list Philosophy as their first major, then 40% of Philosophy majors in the U.S. are double majors. Unfortunately, the NCES database doesn't allow us to see how many of the people with first majors in Philosophy also had second majors in something else. Forty percent might be too high an estimate, if double majors who have Philosophy as one of their majors disproportionately list Philosophy as their second major. But even if 30%, rather than 40%, of Philosophy majors carry Philosophy along with some other major, that is still a substantial proportion.

Here's another way of looking at the data: 0.3% of students choose Philosophy as a first major, while among those who decide to take a second major, 1.7% choose Philosophy.

Across all 37 top-level categories of major (excluding from analysis the top-level category Philosophy & Religious Studies), only two had a higher percentage of students completing the major as a second major: Foreign Languages, Literatures, and Linguistics (28.5%), and Area, Ethnic, Cultural, Gender and Group Studies (26.2%). Let's call this the Second Major Percentage. In this respect, Philosophy is quite different from two of the other big humanities majors with which Philosophy is often compared: English Language and Literature (7.2%) and History (9.3%). Interestingly, Mathematics, which is often regarded as a very different type of major, ranks fourth, with a Second Major Percentage of 14.4%.

Here's a breakdown among all top-level majors with at least 10,000 completed degrees in the three-year period:

[apologies for the blurry rendering; click to enlarge; correction: the second "Engineering Technologies" should be simply "Engineering"]

Last fall, I presented data showing the precipitous decline in Philosophy majors in the past few years -- from 0.58% of all graduates to 0.39% of all graduates. One hypothesis, which I thought worth considering, is that Philosophy tends disproportionately to rely on double majors, and it is increasingly difficult for students to earn a second major. However, I now think that a decline in second majors can't be the primary explanation for the decline of the Philosophy major. First, although the percentage of students earning a second major has declined somewhat in the period -- from 5.4% in 2009-2010 to 5.1% in 2015-2016 -- that is small compared to the magnitude of the decline in Philosophy, where completions are down by 19% in absolute numbers and about a third in relative percentages. Second, we saw similar declines in English and History, and those disciplines don't appear to be as reliant on students' declaring a second major.

That said, Philosophy does appear to rely heavily on double majors, so we might expect policies that reduce the likelihood of double majoring to disproportionately harm Philosophy programs. Such policies might include increasing the requirements for other popular majors, increasing general education requirements (apart from G.E. requirements in Philosophy, of course), and increasing pressure to complete the degree quickly.

To further explore the question, I divided the colleges and universities into two categories: those with a high rate of double majoring (>= 10% of graduating students have two majors) vs those with a low-to-medium rate of double majoring (< 10%), excluding institutions with < 300 Bachelor's degree recipients over the three-year period. As expected, at institutions where students commonly complete second majors, about three times as high a percentage of students completed Philosophy degrees than at institutions where students less commonly complete second majors: 9240/935419 (1.0%) vs. 14876/4656563 (0.3%; p << .001).

Direction of causation is of course hard to know. These groups of institutions will differ in many other respects too, and that is very likely to be part of the causal story, maybe most of the causal story. And yet, the following fact is suggestive: The four majors that differ most in percentage between the two groups of universities, as measured by the ratio of percentage of students completing the major at the high-double-major schools to the percentage completing the major at the low-double major schools, those are exactly the same four majors that have the highest rate of second majors in the overall dataset.

That's a little abstractly put, so let me give you the breakdown for the highest-ratio majors, so you can see where this is coming from:

Area, Ethnic, Cultural, Gender and Group Studies: 1.5% of majors at high-double-major schools vs. 0.4% at low-double-major schools, ratio 4.1.
Philosophy: 1.0% vs. 0.3%, ratio 3.1
Foreign Languages and Literatures: 3.2% vs. 1.1%, ratio 3.0
Mathematics: 2.8% vs. 1.1%, ratio 2.7

I need a good name for this ratio, but I can't think of one, so let's just call it The Ratio. Of course, the median Ratio is about 1.0. History and English both have Ratios of 1.9, which is substantially above 1.0, but not as high as these other four.

In other words, perhaps unsurprisingly, the four majors that students are disproportionately most likely to declare as second majors are exactly the same four majors that show the greatest difference in completions between schools where lots of students have double majors and schools where few do.

With a few exceptions, most notably Biology (a Second Major Percentage of only 2.7%, but a Ratio of 1.7), the relationship is reasonably tidy. The correlation between the natural log of each major's Ratio with each major's Second Major Percentage is r = .76 (p < .001; excluding majors with < 1000 completions in either group of universities; natural log to improve spread near zero). Here it is as a scatterplot:

[click to enlarge]

So although it seems unlikely that the recent sharp decline in Philosophy majors is due primarily to a decline in the overall proportion of students declaring two majors, it remains plausible that conditions that are good for double-majoring in general may be good for the Philosophy major in particular.

Related posts:

Sharp Declines in Philosophy, History, and Language Majors Since 2010 (Dec 14, 2017)

Philosophy Undergraduate Majors Aren't Very Black, but Neither Are They as White as You Might Have Thought (Dec 21, 2017)

Women Have Been Earning 30-34% of Philosophy BAs in the U.S. Since Approximately Forever* (Dec 8, 2017).

Thursday, February 08, 2018

Is Consciousness Sparse or Abundant? Five Dimensions of Analysis

Consciousness -- that is, "phenomenal" consciousness, the stream of experience, subjective experience, "what-it's-like"-ness -- might be sparse, or it might be abundant. There might be lots of it, or there might be very little. This might be so along at least five different dimensions of analysis.

Entity sparseness vs. entity abundance. Consciousness might be sparse in the sense that few entities in the universe possess it, or it might be abundant in the sense that many entities in the universe possess it. Someone who thinks that consciousness is entity-abundant might think that insects are conscious, and snails, maybe earthworms -- all kinds of entities with sensory systems. They might also think that computers could soon have conscious experiences, if designed in the right way. At the extreme, they might accept panpsychism -- the view that every entity in the universe is conscious, even isolated hydrogen ions in outer space. In contrast, someone who thinks that consciousness is entity-sparse is much more conservative, seeing consciousness on Earth as limited, for example, only to cognitively sophisticated mammals and birds, or in the extreme only to adult human beings with sophisticated self-reflective powers.

State sparseness vs. state abundance. An entity who is conscious might be conscious all the time or only once in a while. Someone who thinks that consciousness is state abundant might think that even when we aren't in REM sleep we have dreams or dreamlike experiences or at least experiences of some sort. They might think that when we're driving absent-mindedly and can't remember a thing, we don't really blank out completely. In contrast, someone who thinks that consciousness is state sparse would hold that we are often not conscious at all. Consciousness might disappear entirely during long periods of dreamless sleep, or during habitual activity, or maybe during "flow" states of skillful activity. Someone who holds to extreme state sparseness might say that we are almost never actually phenomenally conscious, except in rare moments of explicit self-reflection.

Modality sparseness vs. modality abundance. An entity who is currently state conscious might be conscious in lots of modalities at once or in only one or few modalities at a time. For example, a minute ago, as you were reading the previous paragraph, did you have visual experience of the computer screen? Did you also have auditory experience of the hum of the fan in your room (to which I assume you weren't attending)? Did you have tactile experience of your feet in your shoes? Did you hear the words of that paragraph in inner speech? Did you have relevant visual imagery? Were you also simultaneously having emotional experience, a lingering feeling of hunger, etc., etc.? Someone who accepts modality abundance thinks that we have many types or modalities of experience ongoing most of the time when we are conscious. In contrast, someone who accepts modality sparseness thinks that we have only one or a few modalities of experience at any one time -- for example, no experience at all of the feeling of your feet in your shoes when your attention is elsewhere.

Modality width. Within a modality that is currently conscious in an entity at a time, the stream of experience might be broad or it might be narrow. Consider visual experience as you are reading. Do you have visual experience only of a few degrees of visual arc, whatever is near the center of your attention? Or do you have visual experience all the way out to the edge, including the rim of your computer screen, the rims of your glasses, the tip of your nose, almost 180 degrees of arc? Or somewhere in between -- maybe the whole computer screen and its rim, but little beyond that? Someone with a sparse view about visual modal width thinks that visual experience is usually (maybe until you think to attend to the periphery) only of a few degrees of arc. Someone with an abundant view might think that we basically experience the full 180 degrees of the visual field all the time when we have any visual experience at all. Analogous issues arise for other modalities. Do you have auditory experience only of the conversation to which you are attending, or also of the background noise? Is your visual imagery sketchily minimal or full of detail? Is your emotional experience a minimal valence and label or is it a very specifically embodied feeling?

Property sparseness vs. property abundance. Consider visual experience again. You are looking at a tree. According to one view, all you visually experience are low-level properties like color, shape, and orientation. You know it's a tree, but the "treeness" of it isn't part of your visual experience. (Maybe you have a cognitive experience of thinking "tree!" but that's a different modality.) According to another view, your visual experience is not just of simple low-level properties but instead of a wealth of properties. It's part of your visual experience, for example, that it's a tree, and that it looks ancient, and that it invites climbing, and that the leaves look about ready to fall.

Philosophers and consciousness scientists have recently been arguing in all kinds of interesting ways about various of these issues in isolation, but I can't recall seeing a good structuring of the landscape of options here which both captures the many different mix-and-match possibilities here, while also recognizing that all of the "abundant" views have something in common (consciousness is widespread and multifarious; it's easy to generate lots of types of it) and all of the "sparse" views have something in common (consciousness is rare and limited in its forms).

[image source]

Friday, February 02, 2018

Designing AI with Rights, Consciousness, Self-Respect, and Freedom (with Mara Garza)

New paper in draft!


We propose four policies of ethical design of human-grade Artificial Intelligence. Two of our policies are precautionary. Given substantial uncertainty both about ethical theory and about the conditions under which AI would have conscious experiences, we should be cautious in our handling of cases where different moral theories or different theories of consciousness would produce very different ethical recommendations. Two of our policies concern respect and freedom. If we design AI that deserves moral consideration equivalent to that of human beings, that AI should be designed with self-respect and with the freedom to explore values other than those we might selfishly impose. We are especially concerned about the temptation to create human-grade AI pre-installed with the desire to cheerfully sacrifice itself for its creators’ benefit.

Full version here. As always, comments and critical discussion welcome, either by email to me or as comments on this post.

[image source]