Emergent Probability

[Richard Moodey]

I have started this topic in the hope that Mounce will report on some of his criticisms of emergent probability.  This is a feature of Lonergan's thought that I have used rather extensively in my teaching of sociology.
Dick Moodey

[mounce.d]

Thanks Dick, you might find an essay or two in Google Groups from the Lonergan listserv, but I guess I could begin by asking how Lonergan might have viewed the nature of probability.  That might provide a foundation, and perhaps even a recommendation, for why emergent is a useful modifier.

[Richard Moodey]

Mounce,
I have never thought that Lonergan's notion of probability -- the probability of an event -- was any different from that which is taught in any elementary statistics class.  The probability of a toss of an unbiased coin coming up heads is .5, etc.   Is your interpretation of him different?
Dick

[mounce]

Hi Dick, these first steps are delicate.  We have introduced two words with complex meaning, "event" and "probability", but I understand what you are roughly saying that the more you "toss" an unbiased coin the closer you expect to get to an average binary distribution.  If we do this a thousand times, for example, then we might employ a quantitative analysis to propose qualitative features that would help us place a bet on one event, or perhaps a series of events.  I think Lonergan deferred from explaining the nature of probability with Insight, and relied on this inferred sense.  His specific idea about a single event was not-quite-right, but his idea about predicting aggregate events is pretty-good.

Remember,
"Ludwig Boltzman, who spent much of his life studying statistical mechanics, died in 1906, by his own hand. Paul Ehrenfest, carrying on the work, died similarly in 1933. Now it is our turn to study statistical mechanics. Perhaps it will be wise to approach the subject cautiously."  (Opening lines of "States of Matter", by D.L. Goodstein).     

[Richard Moodey]

Hi Mounce,

I believe that your analogy to making a bet on a game of chance is an appropriate move in these delicate first steps.  There is another move that Lonergan associates with statistical heuristic structures, and that is the inverse insight.  What do you think of the idea that asking about the "nature" of probability must lead eventually to the inverse insight that probability does not have a "nature" in the same sense as can be discovered in classical inquiries?

My work as a teacher of research methods in sociology was informed by Lonergan's notion of the complementarity of classical and statistical heuristic structures.  I have been very critical of sociologists who assume that the descriptive techniques of correlation and regression will lead to laws analogous to the classical laws of physics.  This is one of the reasons I have used the much simpler device of contingency tables to explain the theoretical significance of correlation and regression.  Each time that a new independent variable is added as something upon which the dependent variable might be contingent, there is a change that the researcher might learn something about the relative probabilities of the occurrence of a each of the values of the dependent variable.

I have used Whitehead's analysis of "events" as "actual occasions," "prehensions," and "concrescences" to fill out my notion of event.  Perhaps some followers of Lonergan would object to this, but, up to this point, I have not been convinced that there is anythng about Whitehead's analysis that contradicts Lonergan's.  Whitehead's analysis has entered into my understanding of the "empirical residue."  I like Whitehead's claim that it is more accurate to say that an event has both duration and volume than it is to say that an event is "located" in time and space.  For Whitehead, time and space are then constituted by the totality of events, which is to say the totality of durations and volumes.  Events constitute the "bottom line" of concreteness. 

The inverse insights that ground statistical method involve a recognition that there is a certain "opaque" quality of the empirical residue that cannot be grasped by direct insights.  To employ a statistical heuristic structure is to anticipate a different kind of intelligibility than that anticipated by a person who employs a classical heuristic structure.   My criticism of some of my fellow sociologists is that their central model or prototype of a scientific inquiry comes from inquiries that have produced the classical laws of physics.

Best regards,
Dick

[mounce]

Hi Dick, I agree, Lonergan's presentation of the inverse insight is profound.  It opens-up fields of inquiry by setting principled limits, and underlies scientific collaboration as a law of nature (engineers love limits!). 

"There has to be a part in the empirical presentations that is *just given* and that does not correspond to any insight."  On this level, the empirical residue has no explanation. 

"Chance," in Ch VIII, however, "is a residual defect of intelligibility," but if no one knows the nature of chance or probability, then I'm not ready to conclude that it necessarily is a defect.  Newton faced a similar challenge trying to understand mass, for example, and now we don't argue about mass in terms of lengths.  Maybe we'll never know this goddess Fortuna, but you are right that all the correlation and regression studies won't be enough to unravel the Humean problem; if understanding is possible (ha ha) then we will need a genius like Maxwell.

PS - Maybe you can start a thread on space and time.

[Richard Moodey]

Hi Mounce,
ISTM that an inverse insight, like a direct insight or a reflective insight, can be mistaken.   You seem to be suggesting that the inverse insight that denies direct intelligibility to chance could be mistaken, and that you hope that it is.

A thread on time and space might be interesting, but I have much more than I can handle reading the posts on the threads that are already up.

Best regards,
Dick

[Artfulhousing]

Dick,

You write “My criticism of some of my fellow sociologists is that their central model or prototype of a scientific inquiry comes from inquiries that have produced the classical laws of physics.”

My criticism is not so much that sociologists seek to produce classical laws but rather that they don’t know what a classical law is and don’t know what a physicist is doing when they discover a classical law. Sociologists tend to caught up in some form of statistical analysis thinking that when they can generalise their analysis to other populations, they have produced a ‘classical law’. When they are doing this they are generalising place/time associations between elements. Insight is involved here but it is not the insight that gives rise to classical laws.

A classical law, however, is discovered in the creative moment of insight when the intelligibility of something is grasped, when this whole and its elements are systematically related to one another (or, what I call a theory - it is abstract, invariant, universal and normative). The key thing here is the grasp of systematic relationships. In other words, (i) these elements are the significant, essential and relevant ones – the insignificant, unessential and irrelevant elements are thereby excluded; (ii) the elements and their relationships together constitute this something.

Classical laws are answers to what-is-it questions. So a mathematician asks what is circle? and discovers through the creative moment of insight that it is constituted by certain elements which are systematically related to one another, viz. “a locus of coplanar points equidistant from a centre” (Insight 31). In this creative moment, they leave aside as insignificant, unessential and irrelevant the size of the circle, the different planes in which a circle can be drawn, the ‘dimensions’ of the central point and the coplanar points.

In a similar way, the sociologist can ask what is a society? what is sociality? What is an organisation? What is an economy? Etc. and discover the set of systematic relationships that constitute it, i.e. the set of sets of activities and practices that constitute a society – a theory of society etc. They can leave aside the interests, motivations, attitudes, beliefs etc. of social agents (whether individuals, groups or institutions). Unfortunately, it is the occurrence of these and their associations with something else (‘discovered’ through statistical analysis) that tend to be the focus of too many sociologists. (By the way, in my view, Lonergan outlines his theory of society on page 47-52 of Method as he explains the diagram on p.48)

[mounce]

Hi Dick (and Art!), Adrial has done some work on the inverse insight that might be useful.  I think Lonergan only uses it to illustrate a principled limit, but there is a sense in the two types that some operations in the field of falsification are included at the level of judgment when you have the outcome of an inverse insight.

That is a little different from finding a, "defect of intelligibility" given that no one understands the nature of probability.

In any case, it might be easier to consider his dialectic in this regard.  The, "events of determinate character," also underlies his notion of emergent probability and the presentation of history.  I think it is picked up again in Chs. IX and X with the upper and lower blades, wise choosing, and propositions turned into wine (I mean principles - ha ha). 

old joke, "Those are my principles, and if you don't like them, well, I've got others"  Groucho Marx

[Richard Moodey]

Hi Art,
Are you saying that every good answer to a question of the form "What is the nature of x?" yields a classical law?  Your use of the answer to "What is the nature of a circle?" to illustrate classical law suggests to me that you are.  I have always used that example to illustrate the difference between a nominal and a real definition.  But I don't take every real definition to be a classical law.  I generally use the law of gravitation as my example.  The abstract qualities of mass and distance determine the force of gravitational attraction: the attraction between two bodies is proportional to the product of their masses and inversely proportional to the square of the distance between them.  This differs from the definition of a circle, both in that it specifies dynamic relationships among measurable quantities, and in that it fails to provide any insight into the internal "nature" of the attraction, even though it can be measured.

[mounce]

When Newton and his editor Cotes were struggling with the concept of mass, Newton was trying to capture mass in terms of lengths as a way to discredit Descrates.  When measuring the proportional attraction between two bodies, do you measure from the center, the surface or somewhere in-between, and are those lengths important?  They understood mass in the same way we understand probability.

[Richard Moodey]

Hi Mounce,
In response to your question about where the distance between two bodies is measured from, I don't know.  Perhaps someone with a better knowledge of Newtonian physics than I might be able to answer.  As to your final sentence, am I wrong to read it as a bit ironic?
Dick

[mounce]

Ironic, perhaps, but I don't mean to be cynical.  It comes from a quote by  Suppes in 1956 that I think is still valid today, and the author turned it into a statement that would have been valid in Newton's time but is no longer; I continue to test its validity by saying (in a variety of conversations with a variety of people), `no one understands the nature of probability.'

In any case, what do you think about Lonergan's definition of dialectic beginning with, "events of a determinate character,"?

[Richard Moodey]

Hi Mounce,

In my classes, I spend a good bit of time with dialectic -- I believe it is an important heuristic for sociology.  I generally begin with Marxian dialectic, since most contemporary sociologists consider Marx to be one of the "fathers" of sociology.  I then contrast Lonergan's notion with that of Marx, following Lonergan's section on "The Dialectic of Community."  I modify Lonergan's second point, however, because the main kinds of events I focus upon are episodes of social interaction.  In these kinds of events, the "principles" are persons, and there can be more that two participants in an episode of interaction.

I treat dialectic as one of three heuristic structures for thinking about social and cultural change.  The other two are what I call the "embryonic model" (Lonergan's genetic heuristic structure) and the "variation-selection model," which is prominent in classical micro-economic theory, operant conditioning theory, and evolutionary theory.  I disagree with those sociologists (including Talcott Parsons) who use an embryonic model to explain what they call social or cultural "evolution."

This is a long-winded response to your question about what I think of Lonergan's definition of dialectic.  I think his definition in this section of Insight is related to his discussion of dialectic as a functional specialization, but in "The Dialectic of Community" he is talking primarily about a social process, and secondarily about a way of thinking.

Best regards,
Dick

[mounce]

Thanks Dick, Lonergan's use of dialectic is dynamic.  I think the introduction with Insight derives from Fernand Braudel's concept of the longue duree, but in regard to probability what do you explicitly think about, "events of a determinate character,"?