The logician and analytic philosopher Michael
Dummet is a dense writer: never wilfully obscure, but the ideas
have heft, and are tight-packed,
the arguments of the sort for which “Follow me closely here, Watson”
applies.
I read any of his essays sentence by sentence,
possibly moving my lips, and pausing for breath after each; and frequently pen-in little
accent-marks, or separation-signs whose value is taken as slightly weaker than
that of commas, to make antecedently clear, upon re-reading, whích clump of
words belong together, and which individual words are to be taken in other than
the unmarked sense. (As,
just there, whích, corresponding to
Spanish qué versus que. -- The
grouping-mark, in manuscript, is a small downtick between minor
constituents; in these essays, I
often place two spaces between such members rather than one, for the convenience of readers.) For in writing such as his, where
the thoughts are not already familiar to the reader, it is easy to start down
an interpretive “garden path”, whenever some connective might be taken in more
than one way and we ill-fatedly
take it (without taking thought) in a given way (generally, the commoner use of
the word), and so doing provoke a “premature collapse of the semantic wave-packet” (thus foreclosing
the continued interplay of mutually influencing ambiguities), and are obliged
to begin again at the beginning.
Consider then this, from his closely-reasoned
essay “Is Logic Empirical?” (a reply to Putnam’s celebrated paper of the same
title)
…
the assumption is that …
The usual reading of the assumption is that X
would be the same as for one assumes that
X, or X is assumed. In this case, the “is that” is
quite unessential, and the phrase could be rephrased with neither copula nor
relative pronoun: Our assumption: X, where the colon has ostensive
force.
That, however, was not the intended use in the
sentence in question, which we now quote unabridged:
Putnam
remarks of this assumption that it is the usual operationalist
idealisation. That is perfectly
true: but the specifically
operationalist part of the assumption is that to the effect that, for every
property, there exists a test for its possession.
Or, rewritten with diacritics:
Putnam
remarks of this assumption that it is the usual operationalist
idealisation. That is perfectly
true: but the specifically
operationalist part of the assumption
is thát to the effect that, for every
property, there exists a test for its possession.
In this sentence, indeed, there are three
occurences of the orthographic word that, each of which is prosodically and
grammatically distinct, and which indeed would each be translated
differently into French, much as
with our Spanish example above.
(1) The first that
in the second sentence is an anaphoric back-glance at the content of the
proposition stated in the previous sentence. Its rhetorical import, and consequent phonetic treatment,
can differ subtly in strength. It
its weaker use, it does not focus at all on the details of the proposition
anaphorized, serving rather as a sort of blandly dismissive wave of the hand,
and accordingly would be completely unstressed, and indeed could even be
omitted completely: the
first clause could be rephrased
All
very true, but…
The French equivalent is likewise phonetically
reduced (c’), and even omissible:
C’est
bien vrai, mais … / Bien vrai, mais …
In a slightly strengthened form, the phrase would
instead take seriously the
proposition being anaphorized,
conceding the point but only in so far, implicitly contrasting it with
congeneric propositions with someone might likewise put forward but which the
present author might reject; the particle could here be stressed:
Thát
(much) is (indeed) true; but …
And here the French would use the full form of the
particle:
Cela
est vrai ; mais …
or even (now mimicking the fuller rephrasing of the English which we denoted using parentheses:
Ce
propos-là est bien vrai; mais, …
or (in a context in which propos or proposition
would be understood)
Celui-
là / Celle- là …
(2) The thát
which I have graphically stressed
could be replaced by the one; the French here would have celui qui.
(3) The remaining occurrence of that , in to the effect that, presents subtleties of its own. In this idiom, that makes no independent contribution, and would accordingly be
destressed, in uses like “… to the effect that a person….”, where the French
could (antevocalically) reduce que to a mere clitic qu’ : “ … dans le sense qu’une
personne…”.
Here, however, the vicissitudes of syntax, leaving
that at the end of a breath-group,
insures that it should not be so fully destressed; nor could French elide (phonetically or orthographically)
elide the schwa in que :
…
to the effect that, in every case, …
…
avec le résultat que [*qu’], en chaque instance …
French here makes a triple orthographic
distinction where English makes none, not because “ce qui n’est pas logique n’est pas français”, but owing to the
heavy effect that stress-within-the-sentence has had in the evolution of (usually curtailed) French words from their Latin originals. English has a few such instances
phonetically, less likely to be reflected orthographically: as, stressed that, with the vowel of cat,
versus destressed, with but a schwa. Or, consider the very different uses of the complex
word always spelled some. I once participated in an ongoing “Formal Grammar Group” at
the University of Alberta, where we discussed things like quantification and
mass terms, where we invented graphies some
versus s’m to make a stab at making
visual distinctions.
One English case in which differential phonetic
erosion did produce a triplet of
distinct spellings is that of one
> an > a. Here orthography has still not quite
caught up with the phonetic evolution:
depending upon the speaker’s intentions, a may be pronounced
either with a schwa, or with
the vowel in day. -- In German, the parallel evolution
(from number-word to indefinite article) did not likewise result in an
orthographic split: ein is used in both senses. However, a careful writer, analytically
inclined, may supply a written acute accent: eín Grund, wofür ….
‘one reason for which …’
~
Let us take a different example from the same author.
First, try to contextualize the following initial
fragment:
Yet
it is a delicate matter so …
Here is the text in full:
Yet it
is a delicate matter so to describe the connection between premisses and
conclusion as to display clearly the way in which both requirements are
fulfilled.
-- Michael Dummett, Truth and
other enigmas (1978), p. 297;
as typeset.
Which, in our scribally enhanced verion, would come out
Yet it
is a delicate matter só to
describe the connection between premisses and conclusion as to display clearly the way in
which both requirements are fulfilled.
Note that here that the reason that the só receives an accent, is not (as in the
case of the last that analyzed above)
that it finds itself fortuitously at the end of a breath-group, but that it
finds itself separated, by a lengthy complement, from a second particle with which it logically forms a single correlative conjunction (like either … or, or neither … nor): the stress, then, prospectively points (cataphorically
links) to its idiom-mate which is eventually to come, much as the full phonetic
form of either / entweder / ou bien
points forward to an upcoming or /
oder / ou.
The use of this figure in the sentence quoted, which is
characteristic of the author’s thought and style, is absent from spoken
language, as from the written of
all but a sliver of the population:
most people would never use it, and might not quite understand a sentence
like Dummet’s above. Indeed it is characteristic, not merely of the
formal or literary style, but of the philosophical, the analytical.
Here is how the plain man uses these words:
John rigged up the contraption so that it would both do your taxes and
toast bread.
In light of that, you might imagine that the unitary connection bolded above
(and quite unstressed; you could
even pronounce it as a sort of monosyllable: s’th’t) represents
the basic or original idiom, and that this logically and phonetically
tense só … that is a later or professorial
refinement. Yet in fact the versions in which
so has a full illative sense, came first, and were widely used in all sorts of
writing in former times: Só that you mind! = “Just so long as you
pay attention!” or God so ordered the
world that each creature might occupy its proper place in the Chain of Being.
Now, in our “contraption” sentence, the so, although
colloquially quite unstressed, still retains the logical role it fills in God so ordered the world that…; it could be replaced by something
fuller, like in such a way that …
; and French would use relatively
mouth-filling equivalents: de manière à ce que, afin de, tant et si
bien que. In this case, the
modification is logically restrictive:
John constructed the thing using thát method which would insure … But these unstressed contiguous
words can easily fade into a
weakened, non-restrictive meaning:
John blundered up against the house
of cards, so that it toppled.
Here there is no design or intention on John’s part, the way
there was with his contraption or
with the crafty Creator. The
relation between the propositions is no weaker: ‘with the result that’, whether or not that result had been
consciously compassed by the
actor. The sentence could indeed be written with no
intrasentential conjuntion at all, the effect of the triggering accident being
exported to another sentence entirely:
John blundered up against the house
of cards. (What a schlemiel! ) It
toppled.
.
Yet it
is a delicate matter so to describe the connection between premisses and
conclusion as to display clearly the way in which both requirements are
fulfilled.
-- Michael Dummett, Truth and
other enigmas (1978), p. 297;
as typeset.
Which, in our scribally enhanced verion, would come out
Yet it
is a delicate matter só to
describe the connection between premisses and conclusion as to display clearly the way in
which both requirements are fulfilled.
Note that here the só receives an accent, is not (as in the
case of the last that analyzed above)
that it finds itself fortuitously at the end of a breath-group, but that it
finds itself separated, by a lengthy complement, from a second particle with which it logically forms a single correlative conjunction (like either … or, or neither … nor): the stress, then, prospectively points (cataphorically
links) to its idiom-mate which is eventually to come, much as the full phonetic
form of either / entweder / ou bien
points forward to an upcoming or /
oder / ou.
The use of this figure in the sentence quoted, which is
characteristic of the author’s thought and style, is absent from spoken
language, as from the written of
all but a sliver of the population:
most people would never use it, and might not quite understand a sentence
like Dummet’s above. Indeed it is characteristic, not merely of the
formal or literary style, but of the philosophical, the analytical.
Here is how the plain man uses these words:
John rigged up the contraption so that it would both do your taxes and
toast bread.
In light of that, you might imagine that the unitary connection bolded above
(and quite unstressed; you could
even pronounce it as a sort of monosyllable: s’th’t) represents
the basic or original idiom, and that this logically and phonetically
tense só … that is a later or professorial
refinement. Yet in fact the versions in which
so has a full illative sense, came first, and were widely used in all sorts of
writing in former times: Só that you mind! = “Just so long as you
pay attention!” or God so ordered the
world that each creature might occupy its proper place in the Chain of Being.
Now, in our “contraption” sentence, the so, although
colloquially quite unstressed, still retains the logical role it fills in God so ordered the world that…; it could be replaced by something
fuller, like in such a way that …
; and French would use relatively
mouth-filling equivalents: de manière à ce que, afin de, tant et si
bien que. In this case, the
modification is logically restrictive:
John constructed the thing using thát method which would insure … But these unstressed contiguous
words can easily fade into a
weakened, non-restrictive meaning:
John blundered up against the house
of cards, so that it toppled.
Here there is no design or intention on John’s part, the way
there was with his contraption or
with the crafty Creator. The
relation between the propositions is no weaker: ‘with the result that’, whether or not that result had been
consciously compassed by the
actor. The sentence could indeed be written with no
intrasentential conjuntion at all, the effect of the triggering accident being
exported to another sentence entirely:
John blundered up against the house
of cards. (What a schlemiel! ) It
toppled.
~
Our colleague Dr Massey comments:
It is my impression that such logical fine distinctions,
supporting distant correlative constructions, are characteristic of English and German, and even moreso of
Latin; less so of Arabic (though
in other ways that classical language is very rich).
[Postscript]
Dr. Massey, the official Latinist of this site (lavishly compensated out
of our enormous profits selling Platonism Rules! coffee-mugs and t-shirts),
weighs in with a parallel to the two uses of so that noted above. In this case, no phonetic/stress
disambiguation is available, since the equivalent is here is single word,
indeed a monosyllable: UT.
Similar remarks could be made about the classical Arabic Hatta.
Pertinent is, perhaps the distinction between the Latin
Purpose Clause and the Result Clause.
Interestingly, the clauses are identical with
the exception of some element in the Protasis such as "So much, so great,
so" (tantum, tot, etc) in a result clause.
Consider the following.
Caesar urbem intravit ut inimicos interficeret.
Caesar entered the city in order to kill the
enemy.
Here we have a simple purpose clause, Caesar did
one thing in order to do another thing.
But the addition of some "so much, so
great" element in the protasis can render the apodosis to be a result:
Caesar tantos inimicos interfecit ut sanguis in
vias fluxerit.
Caesar killed so many enemies that [as a result]
blood flowed in the streets.
Now, Caesar did not kill because he felt the
streets needed more blood in them. He killed because he was a murderous thug.
But the result of his activities was that the streets flowed with blood.
.
