Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

News A Probable Crisis in English Philology.

  1. Sep 1, 2012 #1
    I am quite troubled by a certain singularity in thought which has crept into our common usage. In general, we strive for precision in our words, and tend to inflate and separate words, broadening vocabulary, manufacturing new separate words when we run across ambiguity. Sometimes several disparate concepts are contained in one word and used for one or the other. Context alone guides us to the intended meaning. Sometimes, we allow ambiguity to persist when the disparate meanings have been deeply rooted in our language for a long time. Such things are the business of philology.

    That confusing preamble having been churned out, I am concerned with our historical use of the word 'probability' and how we have allowed two quite disparate concepts to fuse without a protest - to the clear detriment of clarity of thought.

    I single out English for this example, for that is my principal language. It may be a similar confusion in the other languages as well.

    From the original Latin, "probere" is to prove. It gave rise to the words "probative" and "probable," which were originally much closer in meaning, if not identical. Something which is "probable," in the original sense, is that which can be fleshed out into that which is probative, i.e. actually proving.

    In piddling around with the mathematics of likelihood, and post-hoc methods in exploring empirical association, mathematicians made a grave error in attaching the word "probability" to this study, regrettably wrecking the concept when it left the mathematical field and came into common parlance.

    Now, in the common parlance, "probability" means primarily the empirical likelihood of a certain event occurring, without reference to its deterministic origins; and also, to a lesser degree, the capacity of a conjecture to participate in a rigorous causal chain of proof. It is not incorrect to look at Euclid's proofs, and say that they are 'probably true.' However, that phrase itself has lost its meaning. (e.g. Federal Rules of Evidence (United States) Rule 401)

    My philological question is how long ago the term "probability" came into the mathematical language, and also subsequently into the common English language. Do you also see this contradiction? What are the real-life implications?
    Last edited: Sep 1, 2012
  2. jcsd
  3. Sep 1, 2012 #2
    Stephen, thank you for that information and welcome to Physics Forums.

    Unfortunately the technical world has a less than perfect record in this respect which has resulted in much confusion over the years.

    Not only do some technical words differ substantially in their precise meaning, but sometimes different technical disciplines offer different meaning for the same word. 'Vector' would be a case in point.

    Furthermore, the technical meaning of some words has changed over time, either because of better theory or just usage.

    This usage effect is apparent in everday English for example the word 'gay' now has quite a different meaning from that of 50 years ago.
  4. Sep 1, 2012 #3
    Philology addresses the evolution of languages, while linguistics include things like synchronic analysis of how words are used at any given time. To understand any real-life implications then requires a linguistic analysis and not merely philology. Since there is no well established science of linguistics yet this still falls under the category of philosophy and speculation for the present. However, tremendous advances have been made in the field in the last century and with the equally rapid advances being made in other fields like neurology it is quite likely a science of linguistics will be established sometime this century. Ironically when that does occur I suspect one of the first consequences will the slow venting of eons of accumulated hot air from academia.

    Anyway, "common" English is what linguists often refer to as a natural language as opposed to a technical one like mathematics. Some speculate that natural languages are inherently messy and filled with contradictions because it promotes rapid evolution of the language, while such messiness would obviously be a problem for a technical language. It might make sense in the confines of a laboratory to have rigid definitions for terms, but in everyday life people frequently have to "wing it" so speak because it just isn't practical to demand rigid definitions. Furthermore, natural and technical language can evolve together as well. A technical word can make it's way into a natural language, evolve rapidly, and then it's new definition finds its way back into the technical language. You could think of them as just different tools, one a sort of general purpose multi-tool and technical languages more like precision tool sets.
    Last edited: Sep 1, 2012
  5. Sep 2, 2012 #4
    Wuhelion sees the area in which I am struggling, and thanks. I think my observation is too rudimentary to merit appellations like diachronic analysis, and other fancy stuff. I just see a conceptual pit that we have slid into.

    Studiot well observes that language, especially English, is somewhat like a much messier thing than we might like to be part of. Like politics and economic planning, we think that we can think about things better than we do when we actually think about them. I think.

    I am, however, pretty concerned about the dent that the concept of mathematical probability has made in American civil rights. In the rational origins of law, one saw the constable's dilemma - that one has to arrest a person who is not guilty, by definition of due process. How can you arrest an innocent person?

    It was answered by the concept of reasonable suspicion. If a person was arrested upon the constable's reasonable suspicion - that is, a suspicion of commission of a crime that could be arrived at by another reasonable person - that the constable was innocent of wrongful arrest, even if the suspect was proven innocent or never charged.

    Now, the concept has become muddied, and dangerously so. Long ago, and I suspect YEARS before the mathematical term "probability" was used, English/American law used the idea of 'probable cause' to mean the identification of a reason (cause) that can be probative, i.e. linked by an unbroken chain of certainties given the proof of each element. Now, in law, we are seeing the concept of "probable" mushed together with probability.

    The miring of criminal law in the concept of likelihood inference is the delight of the McMedia, and the sport of infotainment groupies.
    For example, is an Arab guy in Grand Junction, Colorado, who has dynamite in his garage and an "Obama Sucks" bumper sticker - well, is he probably a terrorist? Survey says yes. In modern parlance, talking heads say yes. In traditional law, one must say - what are the reasons for all these peculiarities, and is there a weight of evidence for each that ties to a single malign purpose? What if this guy is a well known Tea Party redneck, a silver miner up in the boonies, and keeps his mining explosives in a garage way out on his own property?

    During the foundation of the US, there was a fair amount of trepidation regarding the substitution of mass inference for actual reason. Unfortunately, we have pulled a linguistic pin when we allowed "probable" into the popular verbal playpen.
  6. Sep 2, 2012 #5


    User Avatar

    Staff: Mentor

    I'm not seeing a problem there. In any case, this is politics so I'm moving it.
  7. Sep 2, 2012 #6
    I rather object to the movement, as it shows a lack of understanding of the topic. The topic is what Galavotti explored it in his book on the Philosophy of Probability, which is reviewed:
    I have heard little political debate about the diachronic linguistics of probability and its confusing effect on legal analysis in the criminal rights of jurisprudence. Perhaps I've missed it because I rarely watch teevee. Could anyone perhaps reference a link that shows a scintilla of politics related to the topic?
  8. Sep 2, 2012 #7


    User Avatar

    Staff: Mentor

    I guess I don't understand the point of this thread, then -- I thought your point was that the linguistic problems with the word "probable" were causing legal problems (erosion of rights) in the US. If that isn't the point, could you please explain what your point is?
  9. Sep 2, 2012 #8


    User Avatar

    Staff: Mentor

    It was originally posted in philosophy, but the OP didn't follow the rules. I should have just locked this after it was posted.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook