h1

Consistency

February 12, 2010

Democracy Now offers a good example of what I was talking about yesterday. (Hat tip to Max.) Actually, the discussion of climate-vs-weather seems to be everywhere. Here’s another example of it, and Yulsman, as usual, is on the case.  The NY Times also adds a nice little bulletin board of bullshit — otherwise known as the “Room for Debate” (never mind that there’s no actual debate) — suggesting that somehow the connection between weather and climate is better understood psychologically. (Seriously? WTF?)

Recall that the issue here is that some have been claiming that the weather on the east coast is “inconsistent with” global warming models. The charge, if accurate, would be damning.

In response to these charges, many have replied that these weather events are perfectly “consistent with” the models.

And there, my little piggies, is the rub.

Here’s an interesting quote from the editors of Democracy Now:

We speak to climate scientist Brenda Ekwurzel of the Union of Concerned Scientists, who argues the extreme weather is in fact a part of global warming.

How are we to understand this? What are we to say?

More after the jump.

Certainly one plausible way to understand it is as asserting that extreme weather events are the kinds of things that are predicted by global warming models. This snowstorm is one such extreme weather event. Therefore, “this storm is consistent with climate change.” Call this S1.

But another plausible way to understand this statement is as reasserting consistency in the face of concerns over inconsistency. Nothing about this storm is inconsistent with CC. There is even room to say that storms like this are a part of global warming. Therefore, “this storm is consistent with climate change.” Call this S2.

S1, we might say, implies that this storm confirms global warming. Confirmation, it is important to understand, this storm cannot provide, so we must reject S1. But in rejecting S1, I think we also reject it as viable interpretation of what is being said.

What is instead being said is S2: that the storm is not sufficient to deny CC on consistency grounds.

Now then, what’s interesting about S2 over S1 is that, even though it may not be the case that S2 is a direct response to any particular claim, it is nevertheless assessable in a context in which consistency arguments are made. Many people appear to believe that some weather events disconfirm CC. Just as being consistent with a model cannot confirm CC, so too can it not disconfirm CC.

It is not true, therefore, that claiming consistency must, in all cases, be qualified with a claim about the disjunction between weather and climate. Weather is, in fact, related to climate, after all. That’s how we establish what a given climate is: by stringing together weather patterns. A single swallow does not a spring make, as Aristotle says. So too with weather and climate.

But here’s where it gets even more interesting. It is not necessarily misleading even if one makes the following very strong claim:

S3: “Weather event W lends further weight and support to the theory of climate change.”

S3 is also not necessarily false, at least if one subscribes to a pragmatic theory of meaning. Why not? For the same reason that S2 is not false. Taken in the broader context, S3 can be used narrowly as an assertion of scientific fact (maybe S3′) or as a response to the implication (and the lingering doubt) that many of us feel in the face of the evidence that climate change is going to seriously affect us (maybe S3”). The point is not psychological — how dumb! — but theoretical. The meaning of the utterance is only ascertainable by appeal to its use. In these cases — shucks, we all have our doubts about the science — lo and behold, our experiences are both consistent with and lend support to the science (which is not immediately perceivable). When we have doubts about the science, what we are doubting is the extent to which the world accords with or is consistent with the scientific view.

In this respect, the fact that a ball drops does not confirm or disconfirm the theory of gravity, but it lends supporting evidence to it. If some object does not drop, it is helpful to have an explanation that confirms consistency. Helium balloons do not repudiate gravity, for instance. It would be silly to suggest that they do. But if someone suggests such a thing, it is perfectly reasonable to reply that they are consistent with the theory of gravity, and that, in some respects, the theory can help predict their behavior.

To finally drive this home, consider instead S4.

S4: “Weather event (W) hammers the final nail in Gore’s coffin.”

This is a very different case. In this case, S4 is employed exclusively to deny or disconfirm the science of CC. That should be clear enough. It is used to charge inconsistency. W is inconsistent with Gore’s theory. (If you need independent confirmation that this is one of the charges, check out the image above. I stumbled on that using google.)

How should one respond to this charge of inconsistency? By explaining either that it is not inconsistent with the theory (saying that it is consistent with the theory); or by taking a stronger position and explaining that it is accounted for (or anticipated by) Gore’s theory, just as we might do with the helium balloon.

Or consider this more charitable interpretation.

S5: “Weather event (W) lends further weight and support to skepticism about climate change.”

S5 can’t lend support to skepticism either, because again, this form of skepticism about climate change is predicated on a consistency charge. One is skeptical about climate change perhaps for a variety of other reasons, but if W is to lend weight and support to that skepticism, it is primarily on consistency grounds.

Think it might have something to do with perspective? Here’s an even more positive version, presumably parallel to S3:

S6: “Weather event (W) lends further weight and support to the theory of global cooling.”

It’s not even clear how to understand this position. Why not? Because there is no non-comparative context in which this might be uttered. Moreover, to my knowledge, nobody of any seriousness is making this claim. If they are, then they are likely doing so primarily in the context of the prevailing view about global warming. Similarly, if they are, then the response to them ought not to be that their conclusions are inconsistent, but rather that there are stronger theories and better explanations than global cooling; and that…su-prise, su-prise, su-prise…climate ain’t weather.

So there’s my response in a not-so-tidy little package: the tables are not handily flipped. If the tables are flipped, then they must be flipped in such a way that attends to the use of the utterance. What matters for the assessment of utterances is not strictly what the statement means (representationally speaking) but what is being said (pragmatically speaking).

I’ll close with this profound thought from Prof. Johnson at Columbia in the NY Times Debate Room:

Both the cancer patient and the snowbound share a tendency to overweigh the immediate concrete evidence in front of them, even when it has little relevance.

No shit. That’s because the stuff in front of us is immediate and concrete. (Incidentally, his title? “When experience trumps data.” Experience is data, yo.)

About these ads

8 comments

  1. People are playing cards. They play many hands. Jane suggests that the dealer has stacked the deck in some way that will lead to more extreme (rare) hands. At one point in the game Joe is dealt a hand with 4 aces.

    S1-4 Aces is an extreme hand. Thus, Joe’s hand is consistent with the deck being stacked.
    S2-People have said that a hand like Joe’s hand is inconsistent with a stacked deck. No, Joe’s hand is consistent with a stacked deck.
    S3- Joe’s hand lends further weight to Jane’s theory of a stacked deck.
    S4- Bill says that Joe’s hand is inconsistent with a stacked deck. No, Joe’s hand is consistent with a stacked deck.
    S5- Joe’s hand lends further weight to the deck not being stacked
    S6- Joe’s hand lends further weight to the notion that cards are instead being removed from the deck

    All this utterance stuff places us in a squirrel –around-the-tree situation.

    To utter that a weather event is “consistent with” particular views of the nature of climate change is an empty statement, because it does not allow one to distinguish among the above 6 statements. The context make make a consistency statement more or less justifiable, but it does not alter its essential emptiness. It is necessarily an inkblot. If the statement is made to suggest “consistent with” implies confirmation of a particular theory, it is willfully misleading.

    Note. Minor quibbles–> S5 could be appropriate as every hand that falls within an expected PDF adds support for a hypothesis that the deck is not stacked. S6 is perfectly reasonable statement for the hand as well.


    • I generally think talk of data being “consistent” with a theory or model is shorthand for saying something more interesting. However, if we take it literally, then I think this way of putting things is unhelpful.

      First, consistency matters (in this context) only when the relationship between theory and evidence is deductive. Suppose that if T is true, then E must be true. Hence, if not-E, then not-T. It would then be something significant to say T is “consistent with” E.

      Second, for shopworn reasons, the relationship between theory and evidence is rarely deductive. So, what we should be asking is this – what is the Pr(E/T)? How likely would our evidence be if our theory is true? Presumably, then the claim of interest is this – Pr(extreme weather/GCC)>> Pr(extreme weather/natural variability). The probability conferred on extreme weather events is much greater on the GCC rather than the natural variability hypothesis. But, strictly speaking, both hypotheses are consistent with the evidence.


    • Roger – In the post that all this seems to center on, Jeff Masters points out that two rare storms hit the northeast this winter

      http://www.wunderground.com/blog/JeffMasters/comment.html?entrynum=1427&tstamp=

      So following your analogy, it’s more like Joe was dealt four aces and then a straight flush on the next hand.


  2. Well, I use ‘utterance’ quasi-technically in contrast to ‘statement’ or ‘claim’, since it invokes the use theory of meaning. The issue is not, in this case, what the statement itself says, but what is meant by how the statement is used. Clearly, the use theory is saddled with issues related to speaker intention, but I think there are strong reasons to believe that it can escape over-reliance on speaker intention.

    And this brings us to your concerns about inkblots. I think you’re wrong about that, unless you want to go that direction with all statements. We rely on our judgment to give us better and worse interpretations of inkblots, as well as utterances. Depending on how they’re used, we can make reasonable inferences about what they mean, yes?


    • By inkblot I mean, what does the statement about the hand tell us about the deck? Nothing. (Or, whatever you want it to!)

      There are of course justifiable statements, such as, Ben, please repeat the following:

      “Joe’s hand is consistent with the deck being stacked.”

      But the bottom line is that in all cases, such a statement provides no useful information about the deck and in many/most contexts has great potential to mislead the listener aas to the characteristics of the deck.

      There are better responses, such as “You simply cannot tell if the deck is stacked by looking at an individual hand”


    • That’s only sometimes a better response. Sometimes the Hearer is presented with a complex or loaded question. If the Speaker says “Won’t you accept that this hand is inconsistent with your theory that the dealer is stacking the deck?” The answer is, naturally, yes or no. It’s correct to say that a good response to that question is that no hand could possibly demonstrate the truth or falsity of the theory, but that, again, isn’t an answer to the question.

      “Have you stopped beating your wife?” presents a similar sort of problem.


  3. People are playing cards. They play many hands. Jane suggests that the dealer has stacked the deck in some way that will lead to more extreme (rare) hands. At one point in the game Joe is dealt a hand with 4 aces.

    Does Jane have any basis for making this suggestion? Have others been getting good (relatively rare) hands, but not as good as Joe’s (relatively rarer) hands? Is Joe the dealer?


  4. Try shaved dice.



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 29 other followers

%d bloggers like this: