Animals, Fractions, and the Interpretive Tyranny of the Senses in the Dictionary

The Mischief Rule, an article on statutory interpretation that I published several years ago, begins this way:

A Tennessee statute imposed duties on railroad engineers. If a railroad engineer found an animal or obstruction on the tracks, the statute required "the alarm whistle to be sounded, and brakes put down, and every possible means employed to stop the train and prevent an accident." But what counted as an "animal" on the tracks? Cows and horses, yes. But what else? Did all the trains in Tennessee have to stop for squirrels?

The stop-the-train case poses difficult questions for some interpretive theories, especially textualism. The text does not identify a stopping point in what counts as an animal. Nor is there a dictionary definition that will include cows but exclude squirrels. Is a textualist interpreter duty bound to say that trains really do have to stop for squirrels?

Note that there is no dictionary definition for animal that will mean "sizable animal," or "animal of the sort you would need to stop a train for." If you are willing to look at context, including the mischief, but only if you can first find some ambiguity, then you will be absolutely flummoxed by the stop-the-train statute. The solution, or so I argue in The Mischief Rule, is that interpreters should consider the mischief as an aspect of context not only after concluding that a statute is ambiguous, but also when deciding whether it is ambiguous in the first place.

I thought of these points when I ran across this passage in an essay by jazz critic Ted Gioia. He notes that Sony recently invested "in Michael Jackson's song catalog at a valuation of $1.2 billion." He then added: "no label would invest even a fraction of that amount in launching new artists."

Now anyone reading that sentence knows exactly what Gioia means. He is not failing to get across his message. Nor is he an inept user of the English language. He isn't making a mistake.

What does Gioia mean by "a fraction"? No music label, he says, "would invest even a fraction of that amount in launching new artists." But that, of course, is exactly what a music label would do: it would invest not $1.2 billion in a new artist, but rather a fraction of that.

So what exactly does Gioia mean by "a fraction"? He does not simply mean a part of a whole. If Sony invested $1 billion in a new artist--a fraction to be sure, 5/6 of $1.2 billion to be exact--that would run counter to Gioia's point. So he means by "a fraction" not just a part, but a small part.

But he also does not mean "any part." If Sony invested $10,000 in a new artist, would that disprove Gioia's point? Not at all. But $10,000 is quite literally a fraction of $1.2 billion.

So what does Gioia mean? He means by "a fraction" (1) a part of the whole (2) that is small and (3) yet non-trivial. It's a small part of the whole that's actually something, not so small that we could say--and every reader would understand--"that's not even a fraction of what I wanted!" In other words, "a fraction" is being used as a concept that is bounded on both ends, a part that is neither a hair's breadth smaller than the whole, but also not just a hair's breadth. A fraction, a real fraction.

Yet even though (most) readers of Gioia's sentence will understand immediately what he means, the sense in which he is using the word is not identified in dictionaries. At least not in the three I checked just now, including the Oxford English Dictionary. Dictionaries sometimes get the first and second aspects of Gioia's usage, but not the third.

So, to draw the moral for legal interpreters: the semantic domain of a word is not coextensive with dictionary senses, and a word can be used in context in a way that makes sense to every reader and yet does not match any identifiable dictionary sense. Any good reader knows a slug is not an animal in the Tennessee statute, and any good reader knows that $10,000 is not even a fraction of $1.2 billion. Legal interpreters need to be more than good readers, but not less.

Comments

[Krayt]

Scientists use the term "finite", where they mean non-zero, as opposed to most people in common speak, meaning not infinite. That reaction takes a finite number of seconds.

[MatthewSlyfield]

.

If a railroad engineer found an animal or obstruction on the tracks, the statute required “the alarm whistle to be sounded, and brakes put down, and every possible means employed to stop the train and prevent an accident.”

Note that there is no dictionary definition for animal that will mean “sizable animal,” or “animal of the sort you would need to stop a train for.”

You don’t give the exact language of the statute, but I would suggest that whatever language is in the statute that you paraphrase as “found” itself provides a limitation.

The animal would have to be large enough for a reasonable assumption that an engineer in the locomotive of a moving train would see it. I would suggest that a squirrel does not qualify.

[Noscitur a sociis]

.

[Michael P]

Are you sure you understood that correctly? I'm just an engineer, but the scientists I work with would say "a finite time" to mean that the thing does eventually finish (i.e. not infinite), they would look weird at anyone who thought any event could actually occur in zero time, and they use "non-zero" to mean not zero (especially in the context of a non-zero bias). The answers at https://www.reddit.com/r/ENGLISH/comments/16d05u7/to_you_does_finite_mean_not_zero_or_not_infinite/, for example, are solidly in the "finite = not infinite" camp -- there are a few dissenters, but they neither make strong arguments nor claim relevant experience.

[Michael 11]

It's very common to use "finite" to mean "not infinitesimal" just as it's common to use it to mean "not infinite". Routine, even.

You should be familiar with that usage from your math classes, if you're an engineer.

[Magister]

Mathematicians in different branches use the same words differently; for some, natural numbers start with 1 and for others they start with 0. A common but not terribly mathematical meaning of finite is not too large or too small for measurement; not infinite (bigger than all the finite numbers) or infinitesimal (greater than zero but less than all the finite numbers). But infinite as the opposite of finite is clearly incompatible with this, and that would be the usual mathematical definition.

[Michael 11]

or infinitesimal (greater than zero but less than all the finite numbers)

Note that you’ve already assumed that zero is not a finite number.

Note also that infinitesimals do not need to be greater than zero; zero itself is an infinitesimal quantity. That’s why it’s excluded by the term “finite” in uses that contrast “finite” with “infinitesimal”.

[Bored Lawyer]

Common language is imprecise, and often uses metaphors or vagueness to make a point.

Statutes, on the other hand, are meant to regulate society, on pain of punishment, either monetary or prison. Greater precision should be expected there.

[Krayt]

You wouldn't need it if prosecutors and lawyers didn't deliberately try to stretch things. There are the people who brought you marijuana residue on walls not as evidence of past smoking or posession, but of posession in and of itself.

God damn, clever you, prosecutor and police!

[NvEric]

It's true that law be precise, however, and most certainly, law is not made by engineering standards, nor by engineers. Lawyers and thus law is very imprecise when compared to engineering, think "blueprints," methods. Law tries, but mostly fails modern demands of precision for it, law, to be enough of a determinant to be wholly final, thus we have and require judges, who are likewise imprecise and fallible to which appeals are, therefore, allowed.

Precision in law is impossible due to each person's understanding of, and use of, language itself.

Consider A1, "Congress shall make no law . . ." Does that not allow other parts of the federal government to circumvent the meaning and the words of A1 ? Maybe, but the other parts of the federal government are not given the POWER to act "free will" but they do anyway, at times. A1 plainly is aimed at Congress and Congress only since only Congress, in theory, makes law. However, the federal system has been, and is, subverted beyond its intended purpose.

Law is to restrain and not to allow, but as all know, or should, actions are taken by people freely and then all await freely for someone to challenge those actions, or not so.

[MatthewSlyfield]

.

If a railroad engineer found an animal or obstruction on the tracks, the statute required “the alarm whistle to be sounded, and brakes put down, and every possible means employed to stop the train and prevent an accident.”

Note that there is no dictionary definition for animal that will mean “sizable animal,” or “animal of the sort you would need to stop a train for.”

You don’t give the exact language of the statute, but I would suggest that whatever language is in the statute that you paraphrase as “found” itself provides a limitation.

The animal would have to be large enough for a reasonable assumption that an engineer in the locomotive of a moving train would see it. I would suggest that a squirrel does not qualify.

[Noscitur a sociis]

I don’t feel like digging up a 120 year old Tennessee codebook to check, but according to the case from which Prof. Bray’s example is drawn, the statute was triggered “when an animal or obstruction appears on the track”. Nashville. & K. R. Co. v. Davis, 78 S.W. 1050, 1050 (Tenn. 1902). So not sure that’s going to resolve it.

[MatthewSlyfield]

“when an animal or obstruction appears on the track”

I still think that implies a limitation that it’s something an engineer on a moving train can be reasonably expected to be able to see.

Can something that can’t be seen be said to have appeared? I would argue no.

[David Nieporent]

Well, it is not possible to stop for something one didn't see. But that in no way implies that if one sees something that someone with less keen eyesight might not have seen, that one doesn't need to stop.

[MatthewSlyfield]

"But that in no way implies that if one sees something that someone with less keen eyesight might not have seen, that one doesn’t need to stop."

True, but short of a body cam specifically tuned to an individual engineer's visual acuity, there's no way to prove what a particular engineer did or didn't see.

That leaves the court resorting to some kind of standard as to what a typical engineer would have seen.

[qlangley]

What's a fraction of $1.2 billion?

Well, $1.6 billion is 4/3 of it.

[Magister]

The meaning is pretty clearly "proper fraction", less than 1, even if the sizable part is implied (by the word "even", perhaps).

[PFSchaffner]

Curiously, the dictionaries (or usage guides) that address ‘fraction’ tend to focus on the opposite problem, not its use to mean ‘substantial portion,’ but its use or misuse to mean ‘trivial portion,’ as in “only a fraction of those who own rowing machines actually use them.” But that is because the qualifier ‘only’ influences the meaning, just as the qualifier ‘even’ (“I wouldn’t bet even a fraction of that…”) influences the meaning of ‘fraction’ in the opposite direction. Some learners’ dictionaries may pay attention this sort of praxis, but most dictionaries do not, and none can do so comprehensively. Lexicographers tend to say (I’ve been writing dictionaries for 35 years) that every instance of every word is different; the most we can do is group the examples into rough and untidy heaps, not gloss every example or pretend to cover them all.

[SRG2]

Side note: it's "interpretative".

The rule is simple: replace "-tion" with "-tive" and make no further changes.

Hence
Prevention => preventive (and not "preventative" as some of my follow Britons may say)
Attention => attentive
Interpretation => interpretative. The word "interpretion" does not exist except in error.)
Nation => native

Thank you for coming to my TED (adj., "TEDious")

[Voize]

condition => conditive
gumption => gumptive
auction => auctive
position => positive
At least the last one is a word.

[SRG2]

But not all "-tion" words have an adjectival form, though "auctive" does exist albeit it's long-lapsed into desuetude, and if one were to use the word "gumptive" I think any educated reader would understand. Indeed, to use "gumptive" may well itself be gumptive.

[Voize]

Don't they?
conditional
gumptious
auctionary
positional

[Soronel Haetir]

How about a more marginal case than a squirrel, such as a medium-size dog (say in the 40-50lb range). And the fun thing about dogs is that they range in size from something you as a railroad engineer probably do care about all the way down to ones you likely don't.

[Michael 11]

So what does Gioia mean? He means by “a fraction” (1) a part of the whole (2) that is small and (3) yet non-trivial. It’s a small part of the whole that’s actually something, not so small that we could say–and every reader would understand–”that’s not even a fraction of what I wanted!” In other words, “a fraction” is being used as a concept that is bounded on both ends, a part that is neither a hair’s breadth smaller than the whole, but also not just a hair’s breadth. A fraction, a real fraction.

Yet even though (most) readers of Gioia’s sentence will understand immediately what he means, the sense in which he is using the word is not identified in dictionaries. At least not in the three I checked just now, including the Oxford English Dictionary. Dictionaries sometimes get the first and second aspects of Gioia’s usage, but not the third.

----

The problem with this analysis is that the “but not too small” idea you identify isn’t part of the meaning of the word, which is why (a) current dictionaries don’t mention it, and (b) future dictionaries also won’t mention it.

The usage you focus on is hyperbolic, not literal, and it is driven by the inclusion of the expression not even.

If you ask me to recommend a meal at Chez Snootí for $50, and I burst into laughter and say “At Chez Snootí, $50 won’t even get you a glass of water,” I am not making a mistake, even if water is complimentary and the restaurant will sell you a basket of bread and butter for $39.95.

But that has nothing to do with a special meaning of the phrase “fifty dollars” that isn’t commonly found in dictionaries. It is the nature of hyperbole.

[adamscales]

The best response here.

[Stephen Lathrop]

So, to draw the moral for legal interpreters: the semantic domain of a word is not coextensive with dictionary senses, and a word can be used in context in a way that makes sense to every reader and yet does not match any identifiable dictionary sense.

That by itself ought to dispose of the sketchy practice of legal reliance on dictionary word definitions taken out of context—or even within any lexicographer's necessarily minuscule contextual scope. Historically useful contextual analysis requires far more than even the OED delivers.

But consider this with regard to historical texts which have survived into the present: the larger context which a would-be textualist actually needs is recreated anew from moment to moment, throughout the passage of history. And that process of recreation is necessarily also a process of obscuration, with prior contexts replaced continuously by more recent ones, down through the ages, from era to era.

Thus, every text surviving from an antique past—without exception—arrives in the present bereft of its original context of creation. And quite likely trailing after it like a tail on a kite an appendage of momentary contexts previously attributed to it during evanescent previous, "presents," now also for the most part forgotten.

Confronted with that problem, what does the typical present-day textualist do? The only thing his historically uninformed skills enable him to do: to interpret the text entirely within the scope of present-minded context with which the textualist has been familiar since childhood. Mindfulness of whatever small drift in context might have been noticed during that interval can be counted on to reassure the modern textualist that he enjoys command over whatever problems might arise.

Readers who sense that creates a sometimes-insuperable problem for the vulgar textualism typically praised and practiced by today's legal community are thus correct. But correct with the exception that application of textualism in present context is at least arguably supportable. Confined to texts created in the here and now, there is less reason to distrust the method.

However, pretense to apply that same method to encompass texts descended from eras notably previous to lived experience must be condemned. It is not that those older contexts are never accessible. They sometimes are, to varying degrees, including to high degrees for some texts which were themselves subjects of intense interest and comment during their creation.

The distinction needed is that kind of historical contextual analysis must be practiced by methods entirely different than mentioned above—the application of present context as applied to present text. Instead, the challenge becomes to infer from historical survivals of all kinds—by making them critique each other, and only by that method—the context which characterized a passage of long-forgotten history at the moment of its creation.

That is a process of inference which requires training. It cannot be learned intuitively, the way present context has been learned.

Members of the pro-textualist legal community typically know nothing of that, and do not care that they do not know. They practice a complacence which calls serious purpose into question, disfigures their advocacy for textualism.