PTL:
A Journal for Descriptive Poetics and Theory of Literature 3
(1978) 315- 325
SEMANTIC
EVALUATION OF GRAMMAR IN POETRY
ROBERT
DE BEAUGRANDE
When
readers process a text, they activate the grammatical systems of the language as
organizational principles for sorting the information of the discourse. Under
normal conditions, it is not the grammatical features of the discourse which are
stored in long-term memory, but rather a large-scale representation of the topic
and purpose expressed in or underlying the text (d. van Dijk, 1976). However,
the processing of a poetic text is characterized by increased focus on the
specific forms of the text, including those of grammar. This effect is
frequently determined by the non-ordinary use of grammar and the lexicon in
poetic texts, such that the automatic processing strategies of the reader do not
function smoothly and must, at the least, be adapted. By revising their
expectations during the process of reading, readers save themselves the mental
effort that would have to be expended if every instance of non-ordinary usage
had to be individually confronted with the expectations based on ordinary discourse.
Yet as a heuristic procedure, such a revision suffers from the weakness of its
inductive base. Non-ordinary items in poetry are not necessarily defined or even
used in determinate contexts at their initial occurrence, as would be the case
in technical writing, but may be scattered throughout the text. It should
therefore be the task of literary scholarship to gather the data about language
use in poetic texts, such that readers at large will be able to process such
texts with the appropriate strategies, instead of relying on an incidental
linear series of non-ordinary occurrences.
The
data should be assembled in a word-index and a framework grammar for a given
text, unless the latter is so short that the data can be surveyed at a glance.
The word-index (see also Petöfi, 1971) should list not only recurrent words,
but stems and prefixes/suffixes, together with any lexical groupings which, by
virtue of frequency or positioning, appear significant. Using the index, lexical
items whose usage appears non-ordinary can be determined by a careful
comparison of available contexts of occurrence, thus eliminating the dependency
upon serial ordering and applying the resulting data throughout the text
equally. The grammar should describe differences between the grammatical usage
in the work and that of ordinary discourse, such that it becomes evident whether
or not a given set of differences can be assigned systemic value. The resulting
parameters would not have the status of transformational rules as suggested by
Manfred Bierwisch (1965), but rather of non-obligatory strategies of expression.
In terms of traditional linguistics, these strategies would be seen as
facultative operations between the "langue" and the
"parole."
One
aspect of poetic use of language that can be meaningfully studied with the data
base sketched above is the semantic value assigned to grammatical items under
the increased focus typical of the reception of poetic texts. Roman Jakobson
(1968:602-603) has made the well-known observation that "any noticeable
reiteration of the same grammatical concept becomes an effective poetic device.
Any unbiased, attentive, exhaustive total description of the selection,
distribution, and interrelation of diverse morphological classes and syntactic
constructions in a given poem surprises the examiner himself by unexpected,
striking symmetries and asymmetries, balanced structures, efficient
accumulation of equivalent forms and salient contrasts." This statement has
been frequently cited in studies of poetic language, and its application
demonstrated by Jakobson himself (see footnote 3 on p. 602 of the same article
for a survey). However, it requires some further clarification. The qualifier
"noticeable" brings up the issue of reader focus versus the focus of
the literary scholar or linguist. As I suggested above, and as Jakobson's own
remarks on the "surprise" of the "examiner" indicate, these
"reiterations of the same grammatical concept" may be unnoticed or at
least non-communicative to the average reader. We recall that Michael Riffaterre
was impelled to abandon the "average reader" (Riffaterre, 1959) in
favour of the "super-reader" or "archilecteur" (Riffaterre,
1971:46). Whatever differences we wish to presume between the two, my proposal
of an ordered lexical and grammatical data base made available to readers at
large should reduce the discrepancy —unless we wish to exclude average
readers from the enjoyment of poetry. A related problem is the fact that
ordinary discourse contains a certain amount of obvious and often obligatory
"reiterations of grammatical concepts" to which no one would think of
assigning any special semantic value. The danger of an exaggerated use of the
Jakobsonian method is built into this aspect, and Roland Posner has remarked how
Jakobson himself goes to certain extremes in his analysis (together with
Levi-Strauss, 1962) of Baudelaire's "Les Chats" (Posner, 1969:43, my
translation):
Formal
categories such as gender, number and rhyme-class are semantically milked
without restraint. The investigators seem to assume that plural morphemes always
co-occur with semantic numerousness. Yet that is not even the case with the
Baudelaire poem [ . .
.] The striving to assign content after the fact to formally defined equivalence classes repeatedly
leads the investigators to uncritical analogies.
Still
another problem is the fact that "the description of a11 combinations
appearing in a text is, from the standpoint of volume, an unrealistic
task," as Jurij Lotman (1972:142, my translation) admits. It therefore
appears necessary to establish restrictive criteria to determine which
recurrences of grammatical items should be assigned semantic value in the first
place. 1 would like to propose for a start the following criteria and illustrate
their application to a specific morpheme class:
1)
The items should be non-obvious and non-ordinary, either in their frequency,
their distribution or their con textual environments. For instance, a series of
adjectives agreeing with a noun or a series of verbs in agreement with a subject
would not qualify, nor would items whose frequency is banal, such as auxiliary
verbs or conjunctions like "and." These items should only be
considered if their contextual placement clearly demands focus, such as the use
of "und" [and] as “das süsse Wörtlein” [sweet little word] Act
II of Richard Wagner's Tristan und
Isolde.
2)
The items should not be dictated by the conventions of poetic discourse, or the
grammatical parameters of ordinary discourse in interaction with such
conventions. For example, if Rilke uses forms like "befrein" [set
free] and "eignes" [one' s own] as opposed to the full forms
"befreien" and "eigenes" for the sake of preserving the
meter, such deviations are hardly significant.
3)
The items should be not only markedly atypical oí ordinary discourse with
respect to frequency, distribution or context, but also capable of assuming
autonomous meaning. For instance, the verb "sein" [to be] occurs many
times in the Duineser Elegien, but only sometimes in the emphatic meaning
of "to exist." In contrast to English, these two meanings
are concentrated on this one common verb, thus allowing this
"semanticizing"; Rilke signals his intention further by having the
special usage in italics.
4)
Under ideal conditions, this type of autonomous meaning should be capable of
significant integration into the overall semantic fIow of the text. There will
be some usages which are not in equivalence or contrast with this flow, but this
aspect should always be given consideration.
Although
these criteria may seem too strong or too weak for certain forms of poetry, I
would like to demonstrate their adequacy for the language use in Rainer Maria
Rilke's Duineser Elegien (Rilke, 1955, I: 683-726, hereafter cited by
number of Elegy and line only). The specific morpheme class under investigation
will bethe verbal prefix.
From
the standpoint of ordinary discourse, not all verbal prefixes in German have
retained the consistent meaning or function they may have once had, but many
have. For example, vor- indicates "before in time or space," über-
indicates "over in position or degree," ent- suggests
removal or escape, and so on. Yet these prefixes occur in many verb compounds
where the basic meaning is not fully or even partially maintained, that is, the
prefixes and stems have formed new Units. In the language of Rilke's later
poetry, we often find verbal compounds used in such a way as to detach the basic
meaning of prefixes and stems from integrated units, without, however, losing
track of the conventional meaning of the whole unit. Not only contextual
placement serves this function, but also the creation of numerous wholly new
compounds, some of which have probably never recurred spontaneously before or
since (discounting references to Rilke). A similar technique is found in the
language of Pietismus [pietism], a poetic movement in eighteenth-century German
poetry chiefly deseribing man's position with respect to God and the motion of
forces between them (ef. Langen, 1968:375ff.). In that poetry as well as the Duino
Elegies, these verbal compounds are even used to add an aspect of motion to
actions or states which in themselves indicate none. For instance, when singing
to the angels, the speaker in the Elegies uses the compounds
"ansingen" [sing to or at] and "aufsingen” [sing upward] (ll 2
and X 2). For special emphasis on motion, there are cases of prefix
accumulation, as in "nieder- und herwartstreten" [step downward,
nearer and hither] (ll 7-8) and "abwarts- und hinantreiben" [drive
downward, away and further] (VI 5-6). The prefix ent- occurs thirty-nine
times as opposed to mit- only twice, a fact which even at first glance
seems related to the themes of departure and isolation so prominent in the text. I thus formed the working hypothesis that
the verbal compound as a grammatical item was a possible instance of
semanticizing, and proceeded to compare distribution, frequency and contexto I
found that distribution and frequency were indeed such as to suggest an
intentional arrangement, even though this arrangement would not necessarily
strike the average reader as significant. The comparison of contexts of
occurrence confmned my original hypothesis, though not in all instances.
A
fairly obvious example can be found in the lines describing happiness as
dieser
voreilige Vorteil eines nahen Verlusts (IX 8)
[this
too-hasty advantage of nearing loss]
The
context indicates a supplementary reading of "Vor-teil" as
"advance part," concurring with the statement made at various points
in the Elegies about humans being too hasty in their goals and strivings.
This interpretation is further suggested by the proximity of the "nearing
loss." Thus prefix and word stem have been "semanticized" in the
manner sketched out above.
It
is likewise not difficult to perceive the significance of the prefix auf
[up] in the final line of this passage:
Den
sie von weitem erkennt, ihren Jüngling, was weiss er
selbst
von dem Herren der Lust, der aus dem Einsamen oft,
ehe
das Madchen noch linderte, oft auch als wäre sie nicht,
ach,
von welchem Unkenntlichen triefend, das Gotthaupt
aufhob,
aufrufend die Nacht zu unendlichem Aufruhr. (nI 3-7)
[He whom she knows at a
distance, her youth, what does he know
himself of the master of
desire, who out of the lonely man
(before the girl ever
soothed, often also as if she didn't exist)
uplifted his godly head,
dripping alas with what unrecognizable things,
stirring up the night to
unending uproar?]
A
concentration of three occurrences in a single line can hardly be overlooked,
and the whole passage points to the upward motion in these compounds, although
the prefix only marginally indicates sueh motion in the normal usage of
"aufrufen" and "Aufruhr." It is easier to overlook the fact
that the two forces opposed to this "uproar," namely the maiden and
the mother, are introduced with compounds using the same prefix. Both are said
to have a "soothing" effect (cf. "lindern" [to soothe] in
III 5 and 38); the effect of the maiden begins with her "Ieichter
Auftritt" [lilting arrival] (18), that of the mother with her "zartlich
[...] Aufstehn"
[gentle arising, Le., getting up at night to comfort the child] (39). We note in
line 5 that the primeval uproar was created before the maiden's arrival, but in
the final lines of the same Elegy, the maiden is told that she unwittingly
stirred up "Vorzeit" [primordial times] (77) in her lover. The
synthesis of these two opposed functions assigned to the maiden is further
emphasized by more auf- compounds: "Welche Gefühle wühlten herauf'
[what feelings burrowed up] and "Was für finstere Manner regtest du auf im
Geäder des Jünglings" [what sinister men did you stir up in the veins of
the youth] in III 77-80. Throughout the Third Elegy, a spatial arrangement is
sustained with the primordial elements being low down, viz. such terms as
"Wurzeln" [roots] (51), "stieg er hinab" [he descended] (58)
or "uns im Grunde beruhn" [repose in our foundations] (77). We can
thus conclude that the use of auf- compounds is significant not only for
the immediate contexts of occurrence, but for the thematic development of the
entire Elegy.
The
central topic of the youthful dead is introduced in the First Elegy, and what is
said about them seems unrelated to the rest of the Elegy until the last strophe,
where the human reaction to their fate is seen as bringing “blissful
advancement" (I 90), such as music. The preceding passage (1 69-86) is
summed up as "man entwöhnt sich des Irdischen sanft" [one gently
withdraws from the earthly] (87). The last strophe has no fewer than five ent-
compounds, including that just cited: the dead are termed "Früheentrückten"
[those removed early] (86), a compound unlikely to occur anywhere else at any
time (except in reference to this occurrence); the process of
"withdrawing" cited in line 87 is likened to "outgrowing"
(entwachsen) the mother (87-88); one youth who died is said to have
"stepped out" (enttreten) (94) of a space, another highly non-ordinary
compound; and "blissful advancement" is seen as
"originating" (entspringen) (90) out of human sorrow, in which the
component elements of "entspringen" are semanticized with respect to
normal usage.
These
various circumlocutions for death have been selected in coordination with the
treatment of the topic throughout the whole work, notably in the Tenth Elegy.
There as well as in the First Elegy (ef. I 80-85) it is asserted and
demonstrated that life and death are to be interpreted as similar rather than
opposed in character. This equivalence is supported by using ent- compounds
for life processes as well as death. In the Second Elegy, life is described as
gradual dissemination and dissolution, as in: "Denn wir, wo wir fühlen,
verflüchtigen" [For we, when we have feelings, dissolve] (n 18). Humans
are set in relation to angels. The being of the angels "streams out"
(entströmen) (16 and 31), while that of humans only "goes off” or
"escapes" (entgehen) (28 and 65). Following numerous metaphors of
dissolving (18-36), an exception is apparently made of the lovers, because their
"caresses preserve" them (55ff.). But reservations are expressed. The
statement form is replaced by the question, and soon the whole assertion about
lovers being excepted is undermined. This is underscored by the occurrence of
the same ent- compound (entgehen), in the last line of the strophe (65),
that was used to describe people in general: "o wie entgeht dann der
Trinkende seltsam der Handlung" [o how strangely the drinkers elude their
own action]. In view of this striking instance of morpheme semanticizing, one
wonders if the image in line 50 may be intended as subtIy ironic, opposing the
basic meaning of ent- as used throughout the Second Elegy to an image
portraying just the opposite: "Ihr aber, die ihr im Entzücken des anderen
zunehmt" [You who increase in each other's bliss]. If this were the case,
we would have an instance of a grammatical element being semanticized against,
rather than in agreement with, its immediate context, albeit in accordance with
the overall context. This would clearly be an anomalous manifestation for
semantic theory.
The
prefix hin- is similar in meaning to ent-, but indicates more
decisive and directed motion away from something. In the Sixth Elegy, the
youthful dead are seen as represented by, or at least "wondrously
near" to the hero (VI 21). Since the youthful dead are not mentioned again
in the entire Elegy, one is puzzled as to what the similarity is. The solution
is provided or at least suggested by the hero's motion, namely away. In keeping
with the more energetic and purposeful nature of the hero, hin- compounds
are used rather than those with ent-. The Elegy begins, however, by
contrasting not heroes and the dead, but a fig tree with humans in general. The
tree is called "meaningful" (VII) because of its swift actions:
Wie
der Fontäne Rohr treibt dein gebognes Gezweig
abwärts
den Saft und hinan: und er springt aus dem Schlaf,
fast
nicht erwachend, ins Glück seiner süssesten Leistung. (VI 5- 7)
[Like
the pipe to a fountain, your angular branches drive
the
sap downward and onward: and it springs out of sleep,
hardly
awakened, into the joy of its sweetest achievement.]
An
indicator of intense motion is the piling up of prefixes in line 4. This trait
of rapid energy is set in sharp contrast to the activities of humans, which are
made parallel by using plant metaphors. The parallelism is stressed by use of
another hin- compound, but just before it occurs, there are three
compounds with the prefix ver-, which often indicates that the action
expressed in the verb stem following is done in a mistaken or false manner. That
interpretation is clearly applicable here:
Wir aber verweilen,
ach,
uns rühmt es zu blühn, und ins verspatete Innre
unserer
endlichen Frucht gehn wir verraten hinein. (VI 9-11)
[But
we humans tarry,
alas,
we proclaim our blossoming and enter betrayed
the
retarded interior of our eventual fruit.]
das
uns finster verschweigt, das plötzlich begeisterte Schicksal
singt
ihn hinein in den Sturm seiner aufrauschenden Welt (VI 25-26)
[destiny,
so balefully silent regarding us, with sudden enthusiasm
sings
him into the storm of his soaringIy tumultuous world.]
The
effect of this hin- compound rests not only on its place in a parallel
series, but on containing a stem which in normal usage cannot express any such
motion. Another "foregrounding" technique (as Mukarovsky [1964] might
have called it) is to leave this normally separable prefix attached to a finite
usage of the verb, as in line 43 of the same Elegy:
Denn
hinstürmte der Held durch Aufenthalte der Liebe,
jeder
hob ihn hinaus, jeder ihn meinende Herzschlag (VI 43-44)
[For
when the hero stormed through sojoums of love,
every
heartheat directed to him lifted him hence]
In
the following line (44), we notice still another compound of the series, once
again an action applied to the hero rather than done by him, as in line 26. In
contrast to the compound in line 43, we have hinaus- in 44, but this too
has been prepared in advance by a series of compounds with aus- [out]
placed between members of the hin- series, like the interpolation of the ver-
series earlier, but this time adding not so much a contrast as a
modification. We are told that the hero "broke out" (ausbrechen) (37)
of his mother's body, and that his course was determined by his own
"selection" (Auswahl) (34) and "omission" (auslassen) (36);
here alsó, the prefix in these compounds does not usually indicate motion, at
least not of this type.
The
Seventh and Ninth Elegies, closely related to each other in tope, discuss the
concept of things fading away (cf. in particular IX 11-17). But the tone is more
optimistic than that of some earlier Elegies, since the things are not seen as
dissolving into emptiness (as was apparently the case with humans in the Second
Elegy), but as being "transfigured" (verwandeln) (Vll 49, 51, IX 66,
71). A fresh series of ent- compounds is positioned at key points where
this new interpretation is presented. At the very outset of the Seventh Elegy,
the speaker's voice is declared no longer suitable for courtship, which it
has" outgrown" (entwachsen) (VII 1). While this same verb indicated a
withdrawing from the "earthly" in I 88, it here suggests moving out to
larger functions, such as in fact are subsumed in the Ninth Elegy under the
prominent concept of "sagen" (ef. IX 32-36): This corresponds to a
series of passages in which ent- compounds are used, but at once
qualified or even negated. The maidens only "seemed deprived"
(scheinbar entbehrtet) (VII 40), but in fact did possess complete
"existence" (VII 39-45). Similarly, the plight of the
"disinherited" (Enterbte) (VII 63) need not "confuse us,"
but rather strengthen us in transfiguring and preserving things (VII 65-67). The
interpretation of "entwachsen" I suggested for VII 1 is supported by
this passage:
Tun
unter Krusten, die willig zerspringen,
sobald
innen das Handeln entwächst und sich anders begrenzt
(IX
47-48)
[Activity
under crusts that willingly burst
as
soon as action grows out from within and forms other borders]
The
escape of things into a different type oí being is also suggested by these
lines (note the positioning oí the ent- compounds):
wie
selbst das klagende Leid rein zur Gestalt sich entschliesst,
dient
als ein Ding, oder stirbt in ein Ding —, und jenseits
selig
der Geige entgeht. (IX 61-63)
[how
even lamenting sufferance purely decides to take shape,
serves
as a thing, or perishes into a thing -, and on the other side
blissfully
exudes from a violin.]
Both
the prefix and the stem of "entschliessen" [decide] are semanticized
to allow a reading of "unlocks itself." The same verb returns just
when the speaker declares his willingness to undertake the task of
"transfiguring" (IX 68- 77):
Namenlos
bin ich zu dir entschlossen, von weit her. (IX 75)
[Nameless,
I have decided in favor of you, from far away.]
Thus
the Ninth Elegy closes with a reversal of the original notion that things are
fading away, suggesting that they are in fact — in
"our" internal transfiguration - rather becoming superabundant: "überzahliges
Dasein entspringt mir im Herzen" [overabundant existence originates in my
heart] (IX 79-80); the ent- compound is also to be read as "springs
out." This shift away from loss toward abundance has been further prepared
by a series of über- compounds in the Seventh and Ninth Elegies. The
maidens outdid their lovers and "passed them by" (überholen) (Vll
37). The human achievement of music, which, as we remember, originally was
derived from the loss of a youth (1 91-95), is said to "overreach" (übersteigen)
(Vn 83) its creators. The gaze of humans is called "overfilled" (überfullt)
(IX 20), recalling the assertion in the Eighth Elegy that existence
"overfills" humans (überfüllen) (VIll 68).
The
data assembled above should demonstrate how the restrictive criteria 1 proposed
for evaluating semanticized grammatical items can be applied to the verbal
prefix, and how this method in fact provides far-reaching insights into the
structure of a whole work. The complex distribution of the data in the original
work makes it plain that the gathering and interpreting of such data is a
valuable and rewarding task for linguistic poetics.
REFERENCES
BIERWISCH, MANFRED, 1965. "Poetik und
Linguistik," in: Helmut, Kreuzer & Rul Gunzenhäuser, eds. Mathematik
und Dichtung (München: Nymphenburger). 49-66.
VAN
DIJK, TEUN A., 1976. "Macro-Structures
and Cognition." Paper at the 12th Annual Carnegie
Symposium on Cognition, Pittsburgh, Penn. (University of Amsterdam,
mimeo
JAKOBSON,
ROMAN, 1968. "Poetry of Grammar and Grammar of Poetry," Lingua 21:
597-609.
JAKOBSON,
ROMAN & CLAUDE LÉVI-STRAUSS, 1962. "'Les Chats' de Charles
Baudelaire," L'Homme, Revue française d'anthropologie, II/1: 5-21.
LANGEN,
AUGUST, 1968. Der Wortschatz des deutschen Pietismus (Tübingen:
Niemeyer).
LOTMAN,
JURIJ, 1972. Die Struktur literarischer Texte, trans., Rolf-Dietrich Keil
(München: Wilhelm Fink).
MUKAROVSKY,JAN, 1964.
"Standard Language and Poetic Language," in: Paul Gárvin ed., A
Prague
SchooI Reader on Aesthetics, Literary Structure, and Style (Washington,
D.C.: Georgetown University Press), 17-30.
PETÖFI,
JÁNOS,. 1971. "Probleme der ko-textuellen Analyse von Texten," in:
Jens Ihwe ed., Literaturwissenschaft und Linguistik: Ergebnisse und
Perspektiven 1 (Frankfurt: Athenäum) 173-212.
POSNER,
ROLAND. 1969. "Strukturalismus in der
Gedichtinterpretation. Textdiskription und
Rezeptionsanalyse am Beispiel von Baudelaire's
‘Les Chats’ " Sprache im technischen Zeitalter 29: 27-58.
RIFFATERRE.
MICHAEL, 1959. "Criteria for Style Analysis," Word, 15: 154-74.
RIFFATERRE. MICHAEL, 1971
Essais de stylistique structurale, ed. and trans. Daniel Delas
(Paris: Flammarion).
RILKE, RAINER MARlA. 1955. Duineser
Elegien, in: Sämtliche Werke I [Herausgegeben vom
Rilke-Archiv in Verbindung mit Ruth Sieber-Rilke, besorgt von Ernst Zinn]
(Frankfurt: Insel) 683-726.