96 ANASTASIA M U K H A N O V A KARLSSON
Mukhanov, Ivan Luk janovic. 1999. 'Implicitnye smysly kak sostavnaja cast'
semantiko-pragmaticeskogo potenciala vyskazyvanija'. In E. G. Borisova
& J. S. Martemjanov (eds.), Implicitnost' vjazyke i reci, 81-87. Moskva:
Jazyki russkoj kulitury.
Nikolaeva, Tat'jana Mixajlovna. 1982. Semantika akcentnogo vydelenija.
Leningrad: LGU.
Ode, Cecilia. 1989. Russian intonation: a perceptual description. Amsterdam:
Rodopi.
Svetozarova, Natalija Dmitrievna. 1982. Intonacionnaja sistema russkogo
Jazyka. Leningrad: LGU.
Thorsen, Nina. 1983. 'Two issues in the prosody of Standard Danish'. In
Ann Cutler & D. Robert Ladd (eds.), Prosody: models and measurments,
53-66. Berlin: Springer-Verlag.
Lund University, Dept. of Linguistics and Phonetics
Working Papers 53 (2008), 97-115
97
Early uses of the term chinjutsu
Lars Larm
Introduction
This paper focuses on the early indigenous Japanese works on chinjutsu 1^3^.
Chinjutsu is an important key term within the indigenous grammatical
tradition, even though it seems as if the label 'modality' has taken over
almost completely in recent works. In particular, the notion is central to any
discussion of the history of theories of modality in Japan.
In its ordinary sense, chinjutsu means 'statement' and in combination with
the verb suru 'do' it means 'to state' or 'to make a statement'. The meaning
of the word when used as a grammatical term, however, is much more
elusive and difficult to define, due to the fact that it has been used differently
by different authors. A usual translation is 'predication', but there are other
possible alternatives as well, for example 'modality' or 'illocutionary force',
depending on which work is being referred to. A standard definition, if it is
possible to give one, is found in Koike 1997:288, who says that chinjutsu is
"a sign of completing a sentence, which expresses the speaker's or writer's
attitude, judgment, emotion, etc."
Further, Komatsu 2001:467 usefully distinguishes between two main
themes that have been central in the scholarly debate. Firstiy, it has been
discussed what the function of chinjutsu is within the sentence; and secondly,
scholars differ amongst themselves as to where in the sentence the expression
of chinjutsu is located. As for the functional side, one position is that
chinjutsu has to do with sentence formation, that is, it has a unifying or
synthesizing function which has the force of completing or finishing a
sentence. On the other hand, some scholars point to an addressee-oriented
transmissional function, and yet others use the term in a sense which refers to
modality or subjectivity. There is no unity regarding the location of the
chinjutsu expression either; for example, on one view the verbals have
'chinjutsu-force', while on another it is the final particles that carry the
chinjutsu function.
98 L A R S L A R M
In this paper I shall confine the discussion to how the term was used when
it was still in its infancy.' Since it was Yoshio Yamada 1908, 1931, 1936,
1942 who first used chinjutsu as a linguistic term, I start by outlining some
aspects of his theoryI then go on to survey the central ideas of Miyake,
Tokieda, and Mio.
Yamada: chinjutsu and apperception
It was the works by Yamada that sparked off the debates on chinjutsu? A key
expression in Yamada's theory is tookakusayoo l^lEf^JH which might be
translated as 'a mental operation of unification'. The first part of the word,
tookaku, is the Japanese translation of 'apperception' and sayoo means
'action', 'operation', or 'function'. The word 'apperception' was earlier used
by Leibniz, and then later on by Kant, but it seems to be generally accepted
in the literature that Yamada took his cue from Wundt In his using of the
term. According to Wundt, apperception is crucial to human beings since it is
the cognitive function that unifies elements of our thoughts into a coherent
whole. In his discussion of the 'Apperception centre', Wundt 1904 [1902]:
318 notes that "We thus regard apperception as the one elementary process
indispensable to any sort of 'manifestation of intelligence' and, indeed, to the
higher mental functions at large". This can be compared to Yamada's view
that for each idea or thought there is one mental act of unification, and that
"this tookakusayoo is the life of thought" (1942:425, my trans.). The
synthesizing function plays an important role in Yamada's view of sentence
formation; the linguistic unit of a sentence is viewed as the linguistic
expression of a unified thought. In his own wording (1942:425, my trans.):
A sentence consists of words, but if we only look at its outward
appearance there is nothing but a mere collection of words. The reason
that we can call it a sentence is the internally existing power of the
thought.
Let us now go further and take a look at the notion of chinjutsu. In order to
fully understand the meaning of this key term, it is once again necessary to
'For a more complete review of the literature, including a discussion of later developments
and their theoretical implications, the reader is referred to my doctoral thesis (Larm 2(X)6).
have not been able to consult Yamada's earliest book (1908). I found it, but it was in a
poor condition and difficult to use.
^Yamada's theory of grammar, Yamadabunpoo, is still held in high esteem among
Japanese linguists. Yamada was influenced by indigenous sholars such as Fujitani, but he
was also well versed in German and English works on language, logic, and psychology
(e.g., works by Henry Sweet, Johann Christian Heyse, and Wilhelm Wundt).
E A R L Y U S E S O F T H E T E R M CHINJUTSU 99
move beyond Yamada's own theory and examine what influenced him.
Yamada was knowledgeable in pre-Fregean European logic; and in this
connection there is one point of particular importance. A feature of classical
logic, before Frege entered the stage, is that a proposition is viewed as
consisting logically of three parts: the subject, the predicate, and the copula."
For example, in the sentence:
(1) The rose is red.
the terms 'rose' and 'red' are complete terms and when they are joined
together by the copula 'be' a complete proposition is formed. Thus, the
copula can be seen as a verb which is specialized for this coupling function
and has no semantic content above that. Of course, this is how the logical
structure of sentences is viewed, and the surface structure is a different matter
since not all sentences contain a copula. In this connection it is also worth
drawing attention to the approach that verbs with lexical content have a
copulative function in addition to the function of expressing lexical meaning.
For instance, the verb 'run' has the function of synthesizing the proposition
in addition to the function of expressing the lexical meaning of an activity
performed by the subject. On this view, verbals (including the copula) hold a
unique position within the parts of speech, since they are the coupling
devices needed for predication. Lenci 1998:246-47 notes that predication is
"verb-based" in the theories of Aristotle, Abelard, and the Port Royal
grammarians. In the passage below he discusses Abelard's standpoint (Lenci
1998:243):
[...] it is clear that verbs and only verbs are responsible for the formation
of the predication connection inside statements. Only verbs are able to
express the fact that a certain term is affirmed of an individual.
The verb be is for Abelard the verb par excellence, exactiy because it
signifies in a pure and absolute way the copulative function that
distinguishes verbs from other linguistic expressions.
We are now in a position to discuss what Yamada said about the role of
verbal expressions and the copula in predication. Yamada introduces two
terms: shuikannen i f i f t ^ 'primary concepts' and hinikannen
'secondary concepts', and I shall here use a modem example of my own in
order to illustrate these notions. In the following sentence we have the subject
*This tradition dates back to Aristotle, but the copula went out of fashion when Frege
introduced functions. A predicate, in Frege's view, is "unsaturated" and must combine
with an argument in order to produce a proposition (Frege 1997 [1891]: 139).
100 L A R S L A R M
'Ken' as the primary concept followed by the secondary concept sensei
'teacher':5
(2) Ken wa sensei da.
Ken TOP teacher COP
'Ken is a teacher.'
These two concepts are then linked together by the copula da. Hence, what
we have here is an analysis identical to that in traditional European logic, but
applied to the Japanese language. Yamada 1936:677 explicitly says that in
logic the copula is the linguistic expression which as its only function has this
predicative capacity.
This brings us to Yamada's notion of chinjutsu: the copula is said to have
chinjutsu-force., which can be best translated simply as 'predicative force'.
What Yamada seems to mean is that chinjutsu is the linguistic encoding of
the synthesizing activity of the mind, that is, it represents the unification of
primary and secondary concepts. In other words, the notion of chinjutsu in
the sentence is the 'surface' representation of the 'deep' underlying notion of
tookaku 'apperception'.
The next important point to address is Yamada's view on yoogen
'verbals.' In discussing the European tradition above, it was mentioned that
verbs with lexical content can be regarded as having a copulative function in
addition to the function of expressing the lexical meaning. The same position
is adopted by Yamada, who states that it is the chinjutsu-force that is the
essential characteristic of the verbals, even if it is often the case that the
verbals express the attribute of the subject as well. The fact that the verbals
also can contain the lexical meaning referring to an attribute of the subject is
not seen as very important; it is not a verbal-specific feature since, for
example, nouns also have this content.
The conclusion to be drawn from the discussion so far is that Yamada's
notion of the chinjutsu function is strongly influenced by the pre-Fregean
tradition of logic. In fact, it seems to be a rather standard view of predication,
possibly with the addition of Wundt's notion of apperception to the theory.
However, one important point remains to be addressed. The term chinjutsu
has been used by several scholars after Yamada in various senses, and in
more recent works it refers to modality or illocutionary force. How come the
^The abbreviations used in tliis paper are: COMP = complementizer, COP = copula,
EXEV = external evidence, G E N = genitive, INFIN = infinitive, NEG = negation, N O M =
nominative, NPAST = non-past tense, PAST = past tense, QP = question particle, TOP =
topic marker.
E A R L Y U S E S O F T H E T E R M CHINJUTSU 101
use of Yamada's term, which has to do with predication, has changed so
much during the decades? One possible answer is that the step from
predication and unification to modality is not far. In this connection, it is
worth mentioning that Yamada in one place (1942:49) defines the function of
the verbals as that of a unification judgment, and in another place he even
talks about "special kinds" of chinjutsu such as: meirei 'order', kinsei
'prohibition', gimon 'question', kandoo 'impression', and kakui 'confirmation'.
Therefore, modals are possible carriers of the chinjutsu-fxmciion as
well.
Miyake and early Tokieda
In the previous section we saw that the term chinjutsu, as used by Yamada,
referred to the linguistic coding of apperception, and that he regarded
chinjutsu force as the crucial characteristic of verbals. It will also be
remembered that the copula was considered to be the purpose-made linguistic
sign for expressing this synthesizing activity of the mind; and that ordinary
verbs were seen as hosts of both lexical meaning and chinjutsu force.
However, it was also pointed out that Yamada talked about chinjutsu as a
unification judgment, and that this notion of a 'judgment' to some extent
blurs the picture. Moreover, an additional source of confusion, which was
also touched upon in the previous section, is that Yamada in one passage
mentions several kinds of chinjutsu (for example meirei 'order' and kakui
'confirmation'). In addition to these issues, there is one more problematic
point. In one place Yamada 1936:691 remarks that the chinjutsu is
insufficientiy expressed when the verb occurs in an adnominal position, that
is, in a relative clause. This statement can also be misleading, since it might
lead us to believe that what he actually means by the term chinjutsu is some
kind of illocutionary force or modality, rather than a unifying function. This
last point has caused scholarly debate.
The discussions following Yamada start with Miyake 1934, who at this
stage expresses a favourable view of the notion of chinjutsu. Under Miyake's
analysis the lexical meaning of the verb resides in the stem, and the chinjutsu
function is located in the inflections. This is not a revolutionary suggestion,
however, since Yamada seems to be aware of this, though it is not explicitiy
stated in his writings. In fact, when explaining one example sentence,
Yamada points out that the attributive meaning is encoded in the word
utsukushi 'beautiful' (which here has no nonpast inflection) and that the fully
inflected form utsukushi-i 'beautiful' (nonpast form) expresses both lexical
102 L A R S L A R M
meaning and the activity of a judgment (Yamada 1931:58). Nevertheless,
Miyake's contribution is significant, and he suggests a new category of
jutsushi which appears to correspond not only to inflections but also to
auxiliaries and the copula. Final particles are included in this category as
well. Interestingly, he also argues that the markers '!', '?', and '.' can also
fulfil the task of expressing chinjutsu. It is most likely that by this he means
that the concluding judgment can be expressed by intonation, but a notion of
something similar to indirect speech acts is possibly implied as well. His
three examples have been simplified to the one below (Miyake 1934:23):
(3) Yuki ga fiir-u./!/?
snow N O M fall-NPAST
'It will snow.'/ 'It will snow!'/ 'It will snow?'
Thus, in this example, the same sentence is used to form three different
utterances.
Miyake interprets Yamada's notion of chinjutsu as having the function of
a concluding judgment. It is clear that he also considers affirmation, negation,
and the imperative to be chinjutsu expressions, which is in line with his view
that the location of chinjutsu is in inflections and auxiliaries. Judging from
his examples, the provisional suffix -eba and the politeness marker -masu
are bearers oichinjutsu as well.
Some years later, Miyake 1937 discusses Yamada's explanation of the
insufficient chinjutsu function in relative clauses, which was mentioned
above, and he now adopts a more sceptical attitude. Consider one of the
examples he presents, which is taken from Yamada (Miyake 1937:76):
(4) Hito no suma-nu ie
person G E N live-NEG house
'A house in which nobody lives'
In this noun phrase, the relative clause hito no sumanu modifies the head
noun ie 'house'. Yamada's position here is that the primary concept hito
'person' and the secondary concept, which is the lexical meaning of the verb
sumu 'live', have been unified. The chinjutsu function, however, is not
complete since the verb is used to modify the noun. Now, since Yamada
defined chinjutsu force as the linguistic expression of apperception, this is
somewhat confusing. Miyake 1937:77 claims that there is no chinjutsu
function whatsoever in examples like these, which is a natural position for
him to take since chinjutsu, in his conception, is a notion similar to modality
or illocutionary force. However, and this is slightiy peculiar, he also
E A R L Y U S E S O F T H E T E R M CHINJUTSU 103
questions the unification of the primary and secondary concepts in relative
clauses. Consequentiy, this leads him to doubt Yamada's notion of chinjutsu.
Miyake's remarks do not, however, invalidate Yamada's theory. What
Yamada seems to have meant is that verbals have chinjutsu force in the sense
that they unify the proposition, and he probably did not intend the term to
refer to assertion, illocutionary force, or modality. However, illocutionary
force and modality do indeed take scope over other elements in the clause,
which is clearly seen in the overt structure of the predicate in Japanese, and it
can therefore be expected that there is some overlap in the discussions, even
in the writings of Yamada himself. Probably, he used the expression as a
technical term in some cases and more loosely to mean 'statement' in others.^
At exactiy the same time as Miyake's latest article, and in the same
journal, an article by Tokieda on the concept of the sentence appeared.
Tokieda rejects the view that the defining feature of verbals is that they
express chinjutsu force in addition to the lexical content (Tokieda 1937a: 13).
His argument goes as follows. Firstiy, Tokieda agrees that the copula can
unify the sentence, but he also argues that if the copula has chinjutsu force
then the same should also apply to the final particle in (5) and also to the
marker of external evidence, rashii, in (6) (Tokieda 1937a: 13):
(5) Yama wa yuki ka.
mountain TOP snow QP
'Is it snowing in the mountains?'
(6) Soto wa ame rashi-i.
outside TOP rain EXEV-NPAST
'It seems that it is raining outside.'
Tokieda says that the chinjutsu force in these two examples, similarly to
sentences with a copula, is overtly expressed and separated from the
secondary concept.
Secondly, as for verbs and adjectives, Tokieda takes the position that they
are similar to other parts of speech, such as nouns, in that they do not overtly
encode chinjutsu force. For example, a list of nouns can be meaningful even
if there is no linguistic sign of unification, and a copula can sometimes be
omitted after a noun. On these grounds Tokieda argues that the unifying force
can tacitiy reside in nouns as well, and that it is therefore wrong to assume
the chinjutsu force to be a defining property of the verbals.
'1 owe this observation to Onoe 1990:283, who also offers some textual support for this
interpretation.
104 L A R S L A R M
However, Tokieda's remarks do not seem to undermine Yamada's theory.
Let us first consider his two examples. In (5), which lacks a verbal expression,
the question particle ka does take scope over the rest of the sentence
and it attaches directly to the noun, and, it is indeed the case, as we shall also
see later, that final particles conclude sentences. But to agree to this is not the
same thing as saying that they have the unifying function that Yamada had in
mind. Recall that Yamada took the position of pre-Fregean logicians, and that
a proposition in his view logically consists of a primary and a secondary term
which are joined together by the copula. We are thus talking about propositions,
not actual utterances, and Yamada is well aware of the fact that the
coupling device may or may not be linguistically expressed.
We can probably solve this problem simply by appealing to the competence
and performance distinction. That is, a native speaker of Japanese
tacitly knows that the verbal expression is missing in example (5), and we
can therefore claim that it is the verbal expression that unifies the proposition
in the underlying structure.
As for example (6), with the adjectival predicate extension rashii, Tokieda
is right in saying that it has chinjutsu force in the sense Yamada used the
term. However, this by no means contradicts Yamada's theory, since rashii is
adjectival and therefore could be considered a verbal expression.
Tokieda's argument that the chinjutsu force resides in nouns as well as in
verbals is also odd. Although a list of nouns can be interpreted meaningfully
this does not entail that the nouns in themselves carry the chinjutsu function.
The fact that the unification takes place anyway, without any verbal expression,
can also be explained by appealing to the underiying linguistic system.
That is, we can yet again say that it is the verbal expression that logically carries
the chinjutsu force. It seems that Tokieda, at least in this case, is restricted
since he only considers the performative surface structure of language.
It should be stressed, however, that despite these shortcomings Tokieda
also provides a valuable insight. In part two of the article, he notes that if we
consider the structure of Japanese, the function of the copula can be
described as "wrapping up" the proposition rather than linking the primary
and secondary terms (1937b:69). I shall return to this point in the penultimate
section when considering Tokieda's later works.
Mio: the separation of chinjutsu and unification
In the previous section we noted some unclarities in Yamada's theory. Mio
1939 attempts to solve these problems, and demonstrates a profound
E A R L Y U S E S O F T H E T E R M CHINJUTSU 105
understanding of Yamada's writings. He notes that in Yamada's theory
chinjutsu is seen as the overt linguistic expression of apperception, or, put
differently, that "the process of unification is the content of chinjutsu" (1939:
168, my ti-ans.). The major contribution by Mio is that he established that two
different concepts are intermingled in Yamada's usage of the term chinjutsusayoo
'process of chinjutsu' (or 'chinjutsu operation'). This observation
solves the issues raised by Miyake and Tokieda. In order to see what Mio
means, let us return to (4) from the previous section, here repeated as (7):
(7) Hito no suma-nu ie
person N O M live-NEG house
'A house in which nobody lives'
As explained earlier, in this noun phrase the relative clause modifies the head
noun ie 'house'. Mio points out that the primary concept (the subject) hito
'person' and the secondary concept sumu 'live' in this clause indeed are
unified by the process of apperception, and on his interpretation this is also
the position held by Yamada. Still, Mio states that he is also in line with the
position taken by Miyake that the chinjutsu force is non-existent, and not just
insufficient as Yamada said, when the verbal is in an adnominalized position.
Thus, the problem here is that in the above example the operation of
unification has been at work, but there is still no chinjutsu force. Therefore,
the concept of chinjutsu cannot be the same as the mental process of
unification. These two concepts are essentially different and must be kept
apart. To express this in more modem terminology, what Mio makes explicit
here is that Yamada's term chinjutsu should be divided into two separate
functions: 'unification' and 'judgment'.
Let us now look at the details of Mio's theory. In what follows I shall first
present his view on the process of unification and then go on and describe his
view of chinjutsu. This also follows the structure of his article.
Mio shares Yamada's view that the mental operation of unification is the
"most fundamental operation in thought formation" and furthermore that
"there can be no concluded thoughts without the mental act of unification"
(Mio 1939:170, my trans.). He furthermore suggests that this process can be
seen from two altemative perspectives: the expression point of view
(production) or the understanding viewpoint (interpretation). Let us start with
the former and consider Mio's view on how a thought is built up and unified.
Mio states that the unification process is cmcial for thought formation, and
it thus takes place before a sentence is formed. This operation results in a
complete thought which can lay the foundation of a sentence. He notes that
106 L A R S L A R M
"there are no sentence expressions without the preceding unification process"
(Mio 1939:171, my trans.). Furthermore, he argues that concepts like subject
and object are already included at the stage of the unification process, but
only in a rudimentary form. They are differentiated later at the sentence
stage. The theoretical claim that the unification process is at work on the
'thought level' rather than at the 'sentence level' suggests that Mio has an
idea about something that resembles the Fodorian language of thought, or
maybe alternatively a Chomskyan type of deep structure or logical form.
Matters are different when the unification process is considered from an
interpretive and understanding point of view. Then there is no underlying
thought as the starting point, but rather an already-produced sentence. The
process of unification in this case operates on the level of the sentence, and
Mio notes that "it is an active, dynamic operation" and also that "at the same
time as the progression of this operation we gradually get the content" (Mio
1939:172, my trans.). In Mio's view, the object of the study of grammar is
the sentence level, that is, grammar is concerned with the interpretive side.
We now turn to the question of where the unification process takes place.
Mio questions Yamada's view that the unification process is restricted to the
verbals and argues that the words preceding the verb must also be connected
in some way. Thus, the unification process is at work throughout the
sentence. In support of this view, Mio argues that a list of nouns can be
understood as a unified whole even without a verbal expression. It should be
pointed out that he is aware of the fact that it might be claimed that we
covertly add the verbal expression in such cases, but on his view the
unification process must still take place before we add the verbal expression.
In this connection Mio also considers the role of the copula. He compares
the German proposition
(8) S ist P.
with its Japanese counterpart
(9) S wa P da.
and observes tiiat what is expressed by the copula in the former example is
represented in the latter sentence by both the marker wa and the copula da.
Therefore, the proposition in Japanese contains something in addition to the
primary and the secondary concepts, and the copula. Or, alternatively, the
role of the copula is fulfilled by the combination of wa and da. Mio draws the
conclusion that da on its own, without the marker wa, does not function like
E A R L Y U S E S O F T H E T E R M CHINJUTSU 107
the German copula ist. According to Mio this is further evidence for his claim
that the unification operation is not restricted to verbals; sometimes the
verbal expression fulfils the copular function on its own but sometimes the
marker wa occurs as well, and in such cases the unifying function is not
limited to the verbal expression. Mio explains (1939:175, my trans.):
That is to say, in a proposition like "S wa P da" the process of
unification is at work from the beginning to the end, and it is not the
case that the process of unification is restricted to da.
Again, Mio anticipates how his argumentation could be criticised in saying
that one might argue that the function of wa is to contrast the subject and the
predicate, and that it is, after all, the copula da that does the work of
unification. However, he counters this argument by explaining that to say that
there is a contrastive function is tantamount to saying that there is unification.
Mio then goes on to provide a fine-grained exposition of his chinjutsu
theory. He refers to Yamada's expression tooitsuhantei 'unification
judgment', and proposes that the term chinjutsusayoo 'chinjutsu operation'
should only be used in this sense, i.e., to express the notion of a 'judgmental
operation'. As an illustration, he uses the following example (Mio 1939:177):
(10) Nihongun wa yuukan da.
Japanese troops TOP courage COP
'The Japanese troops are courageous.'
The primary concept Nihongun 'Japanese troops' is here connected with the
secondary concept yuukan 'courage', and the wa-da construction unifies the
whole. The process here, which constructs the complete thought, is a lower
level unification process. Mio then provides a theoretically appealing
suggestion: when the lower level unification has taken place then the thought
is unified, but in addition to this a higher level of unification needs to take
place in order to connect the thought to a state of affairs in the real world.
Both the lower level and the higher level unification processes are seen as
relational operations, so we have not yet reached the notion of chinjutsu.
However, when the relation between the thought and a state of affairs in the
real world has been established, then the speaker can make the judgment that
the statement is true. This judgment is the chinjutsu operation, and it operates
on a higher level than the process of unification.
And, yet again, Mio anticipates criticism for his position. He notes that it
might be argued that we do not need to differentiate between the higher level
unification process and the process of a judgment. But he maintains that the
108 L A R S L A R M
two are different. The former process is relational, while the latter is
'intuitive' and works at a higher level.
We can summarise Mio's views in the following manner. The unification
of concepts takes place in some kind of language of thought. This complete
thought in turn needs to be unified with a state of affairs in the real world.
When this connection has been established a judgment can be made. This
judgmental chinjutsu operation is different from the unification processes and
it operates on a higher level. The process of building up an assertive
statement, from the pre-sentence stage to a complete utterance with
illocutionary force, can therefore be described as follows:
Lower level unification process (relational): works on all parts of
the sentence and creates a imified thought.
i
Higher-level unification process (relational): connects the unified
thought with some state of affairs in the real world.
i
The chinjutsu process (judgmental): the speaker asserts that the
thought matches reality.
It is also worth mentioning that Mio hints at the possibility of other
processes operating at the same level as the judgment, for example questions.
Later Tokieda
Before looking at Tokieda's use of the term chinjutsu, let us consider some
general aspects of his theory of grammar. Tokieda was influenced by two
schools of thought: the indigenous Japanese tradition and the European
philosophical school of phenomenology.^ His theory of language is called
gengokateisetsu @"tf i S ^ f j i 'the language as a process hypothesis', which
he describes as follows (Tokieda 1961:18, my trans.):
The language as a process theory stipulates that language is the act of
expressing ideas or the act of understanding. It says that it is the
expressional process itself, or the process of understanding itself, that is
language.
''I first became aware of Tokieda's strong affinity to phenomenology thanks to an
informative paper by Imai 2002 which is devoted to this topic.
E A R L Y U S E S O F T H E T E R M CHINJUTSU 109
Tokieda divides words into the categories of shi Mand ji The former
category consists of words that undergo a conceptual process while the latter
represents words that reflect the linguistic activity of the speaker. That is, shi
are objective words with lexical content and ji are subjective grammatical
words. A sentence is then gradually built up, in layers as it were, by the
successive combination of these two categories. Let us illustrate this with
Tokieda's famous example (Tokieda 1950:248-250):
(II) Ume no hana ga sai-ta.
umetree G E N flower N O M bloom-PAST
'The ume trees have bloomed.'
Each shi element combines with a ji element to produce yet a larger unit,
which then in turn is embedded within the following shi element, and so on.
In tiie example above ume 'ume tree' (shi) merges with the genitive marker
no (ji), and the newly established unit is then embedded by hana 'flower' (ji).
This process goes on until the sentence finally is wrapped up by a constituent
belonging to the ji category, in this case the past tense marker ta.^ Tokieda
labels this design of the sentence ku no irekogatakoozoo "KJ ® Atl^M#it
'the nesting-type-structure of the clause', and he illustrates the previous
example as below (1950:250):
llJme no hana ga sai ta
Recall from the section before last that Tokieda 1937b:69 provided a
different analysis of the copula than that of Yamada. The observations we
have just made explain why Tokieda argues that the function of the copula,
which he classifies as a ji element, is to wrap up the proposition rather than to
link the primary and the secondary terms.
Another characteristic of Tokieda's theory is the use of empty categories.
In the above sentence the need for a zero marker does not arise, since
Tokieda recognizes the past tense marker (ta in the example above) as a ji
element. However, somewhat inconsistentiy, the nonpast tense morpheme is
not given the same treatment, and the sentence:
8The iterative process that Tokieda describes here resembles the Chomskyan concept of
'merge' within the minimalist framework; functional heads merge with lexical categories
and form larger units. Kinsui 1997:145-146 also observes the resemblance with generative
grammar, and draws attention to the notions of 'projection' and 'adjunction'.
110 L A R S L A R M
(12) Inu ga hashi-ru.
dog N O M run-NPAST
'The dog runs.'
is therefore represented as (1950:258):
Inu ga hashiru | J
where the dark area stands for the non-overt ji marker. Furthermore, nonovert
ji elements appear in attributive positions as well, as in the sentence:
(13) Haya-ku nagare-ru.
fast-INFIN flow-NPAST
'(It) flows fast.'
where the zero marker is placed between the adverb and the verb (1950:273):
Hayaku Mnagareru
Tokieda got the idea about shi and ji from traditional Japanese grammar,
where the category corresponding to ji (subjective words) goes under the
labels teniha X or teniwoha T f C ^ t i . The work that particularly
influenced Tokieda was Akira Suzuki's book Gengyoshishuron (1824).^ In
this work, Suzuki proposes four parts of speech of which the grammatical
subjective words teniwoha constitutes one category. Suzuki observed that the
teniwoha group behaves differently from the other three parts of speech (the
shi words) and he says, for example, that the ji words are 'the voice of the
mind', and that they do not have any reference. A grammatical ji word is also
metaphorically seen as the hand that operates the tool, while the shi word is
the tool itself. Suzuki also noted that the content words, shi, cannot function
without the grammatical ji words, and conversely that the ji words cannot
stand alone.
Thus, we see that the shi and ji distinction used by Tokieda has its base in
traditional Japanese grammar. However, this should be seen in connection to
another important aspect of Tokieda's view of language. In a lecture held in
Nagoya 1967, Tokieda (1968:11-12) said that at first he did not understand
what Suzuki meant and what the grounds for the classification of word
classes were. Then the phenomenological school of philosophy came to his
rescue, and in particular it was Husserl who caught Tokieda's attention. Of
'I have not been able to get hold of Suzuki's book. My remarks on his work are solely
based on Tokieda's explanations.
E A R L Y U S E S O F T H E T E R M CHINJUTSU 111
particular interest here are the Husserlian terms noesis and noema, which
pertain to subjectivity and objectivity respectively. As Moore puts it
(1942:81, emphasis in original):
The 'intentional experiences' themselves are further distinguished into
noesis and noema, the subjective and objective aspects of the
experience, act and content, experiencmg and that which is experienced.
Tokieda talks about ji words and shi words in terms of noesis and noema
expressions, where noesis stands for the intentional act and noema for the
objective side (Tokieda 1968:23). Thus, Tokieda's theory of subjective and
objective elements in the sentence is an application of Husserl's theory of
intentionality. Shi elements stand for the objects of our intentions, and the
intentional act of the speaker manifests itself overtly in ji elements.
Let us now ask the question why Tokieda's theory is significant. If we see
the shi-ji dichotomy as a distinction between function words and content
words then it is of course not a novel idea. The Aristotelian labels
'syncategoremata' and 'categoremata' were widely used in medieval
European logic, and we have seen that the same theoretical distinction has
long been noted in the indigenous tradition as well. What is important,
however, is Tokieda's semantic description of the difference between these
two groups of words. It was mentioned above that shi elements undergo a
conceptual process while ji elements reflect the linguistic activity of the
speaker. In this coimection, we once again need to consult Husserl. In fact,
even if nowhere explicitiy stated, it is not unreasonable to presume that
Tokieda was inspired by Husserl's ideas of 'independent meanings' and
'dependent meanings' which are described in volume two of his Logical
investigations (Husseri 2001 [1901]). Husseri says (2001 [1901]:55):
We must not merely distinguish between categorematic and syncategorematic
expressions but also between categorematic and syncategorematic
meanings. It is more significant to speak of independent and
non-independent meanings.
The point here is that Tokieda is doing exactiy what Husserl considered
important, namely, he analyses the shi and ji in terms of meaning.
This brings us to Tokieda's view of chinjutsu. The key point here is that
Tokieda uses the term chinjutsu to refer to the function of the ji elements in a
sentence (1941:334). It follows that he uses the term not only when referring
to the function of the final ji elements that concludes the sentence, but also
112
L A R S L A R M
for other elements, for example the genitive marker no as in (Tokieda 1941:
244):
(14) Haru no ame
spring G E N rain
"The rain of spring'
The function of the genitive marker in this example is referred to as sooteiteki
chinjutsu ^^6^ 'junction chinjutsu', which can be contrasted with the
notion of jutsugoteki chinjutsu 3ilfMj|5^3# 'nexus chinjutsu' of the ji
element concluding a sentence.
An attractive side of this theory is the use of one general principle of
merging to explain both the unification of the various parts of the sentence
and the wrapping up of the sentence as a whole. If we take the shi and ji
distinction to represent the difference between functional (relational) and
lexical categories, and if we assume that the continuing merging of these two
is the basic principle in sentence formation, then Tokieda's minimal approach
is useful. Nothing hinders us from using the term chinjutsu in this way, if our
aim is restricted to explain the unification process. Of course, the theory may
have some problems, for instance the inconsistency I mentioned earlier when
discussing the empty categories, but this is only to be expected of a
framework still in its infancy.
But, the weak point of Tokieda's theory is the sweeping assumption that//
words encode the activity of the speaker. This overgeneralisation fails to take
into account that there are different kinds of functional categories. We do not
want a tiieory where, for example, case markers are analysed as being
subjective.
Tokieda's grammar is not a theory about chinjutsu in the sense Yamada
used the term. Nevertheless, Tokieda is important because, as Nitta 1988:156
also points out, with him the chinjutsu debate came to be concerned with
subjectivity. In addition, his idea about the layered structure of the clause was
well before its time. The scholarly discussions following Tokieda provide
much useful insight into the nature of subjective modality.
Conclusion
We have seen that the chinjutsu debate was started by Yamada. He drew his
inspiration partiy from traditional logic, where a proposition is viewed as
consisting of a primary and a secondary concept, which are being linked
together by the copula, and partly from Wundt's psychology. Thus, chinjutsu
E A R L Y U S E S O F T H E T E R M CHINJUTSU 113
is defined as the overt linguistic expression of the synthesizing activity of the
mind (tookakusayoo 'apperception').
A source of confusion, however, is that Yamada seems to be ambiguous in
his use of the word chinjutsu. The term is sometimes used in a technical
sense, as it was defined above, and at times in a looser, non-technical way,
meaning 'statement'. This can probably explain why he also talks about
chinjutsu in terms of 'judgments' and other modal notions such as 'orders'
and 'prohibitions'. And, in view of this, it is understandable that Yamada said
that the chinjutsu-fome is insufficient in relative clauses. What he probably
meant was that there is no assertive force in such examples, which is indeed
correct.
Miyake, Tokieda, and Mio defined their positions in relation to Yamada.
Miyake's analysis concentrates on the latter aspect of Yamada's theory, and
accordingly he views chinjutsu as illocutionary force or assertion. However,
he sees a problem in Yamada's seemingly contiradictory statement about the
insufiicient chinjutsu-force in relative clauses.
Both the earlier and the later writings of Tokieda convey a somewhat
confused understanding of what Yamada meant by chinjutsu, but it should
also be recognised that he puts forward some powerful ideas. His main
insight is that the structure of the clause is layered. Another appealing aspect
of his theory is that one general principle of merging is used to explain both
the unification of elements within the sentence and the wrapping up of the
sentence as a whole. Furthermore, Tokieda has had a profound and lasting
influence since he directed the attention to subjectivity.
Although the works of Yamada, Miyake, and Tokieda are still worth
reading (in particular Yamada), they are not entirely satisfactory. Mio, on the
other hand, provides a persuasive altemative. While demonstrating a
thorough understanding of Yamada's work, he illuminates the situation by
drawing a clear and well-motivated distinction between 'unification' and
'judgment'. The word chinjutsu, then, is confined to the latter notion.
Mio anticipated ideas such as 'the language of thought' (or 'logical
form'), and his explanation of chinjutsu as a judgmental operation seems to
resonate well with trath-conditional semantics.
References
Frege, G. 1997 [1891]. 'Function and concept' [Trans. M . Beaney]. In M .
Beaney (ed.). The Frege reader, 130-148. Oxford: Blackwell.
114 L A R S L A R M
Hattori, S. et al. (eds.)- 1978. Nihon no gengogaku [Linguistics in Japan] Vol.
3. Tokyo: Taishuukan.
Husserl, E. 2001 [1900/1901]. Logical investigations V o l . 2. [Trans. N .
Findlay]. London: Routledge.
Imai, R. 2002. 'Tokieda Motoki no kokugokyooikuron: genshoogakuteki
'shutai'-ron kara' [Motoki Tokieda's theory of national language
education; from the point of view of the phenomenological 'subject'
theory]. Paper presented at the 23rd monthly meeting of the
Kokugokyooikushigakkai [The Society for historical studies of teaching
Japanese as a first language], Waseda University, Tokyo. [Available at
http://www8.ocn.ne.jp/~nyororo/academic/gensyogaku.html]
Kinsui, S. 1997. 'Kokubunpoo' [Traditional Japanese grammar]. In T.
Masuoka et al., Iwanamikooza gengo no kagaku 5: bunpoo [The Iwanami
course on the science of language 5], 119-157. Tokyo: Iwanami shoten.
Koike, S. 1997. 'Chinjutsu' [Predication]. In S. Koike et al. (eds.),
Nihongogaku kiiwaado jiten [A dictionary of key terms in Japanese
linguistics], p. 288. Tokyo: Asakurashoten.
Komatsu, K. 2001. 'Chinjutsu' [Predication]. In A . Yamaguchi & S. Akimoto
(eds.), Nihongo bunpoo daijiten [A comprehensive dictionary of Japanese
grammar], 467-468. Tokyo: Meijishoin.
Larm, L. 2006. Modality in Japanese. D.Phil, thesis. University of Oxford.
Lenci, A . 1998. 'The structure of predication'. 5y«f/zej^ 114, 233-276.
Mio, I. 1939. Bun ni okeru chinjutsusayoo to wa nanizoya [What is the
chinjutsu operation in sentences?]. Kokugo to kokubungaku [National
language and literature] 16.1. Reprinted in Hattori et al. 1978,166-181.
Miyake, T. 1934. Onseikoogohoo [Grammar of the spoken language]. Tokyo:
Meijishoin.
Miyake, T. 1937. 'Dooshi no rentaikei ni kansuru hitotsu no utagai nitsuite'
[About one question relating to the adnominal form of verbs]. Kokugo to
kokubungaku [National language and literature] 14:11,70-79.
Moore, J. S. 1942. 'Is this phenomenology?' Philosophy and phenomenological
research 3:1, 78-84
Nitta, Y . 1988. Bunpooronsoo [The grammar debate]. In H . Kindaichi, O.
Hayashi & T. Shibata (eds.), Nihongohyakkajiten [An encyclopaedia of
the Japanese language], 152-162. Tokyo: Taishukan.
Onoe, K . 1990. 'Bunpooron: chinjutsuron no tanjoo to shuuen' [A historical
survey of grammatical theories: the rise and fall of chinjutsu-ron]. Kokugo
to kokubungaku [National language and literature] 67:5. Reprinted in K .
E A R L Y U S E S O F T H E T E R M CHINJUTSU 115
Onoe, 2001, Bunpoo to imi [Grammar and meaning], vol. 1, 277-314.
Tokyo: Kuroshio shuppan.
Tokieda, M . 1937a. 'Bun no gainen nitsute' [On the concept of the sentence]
Part 1. Kokugo to kokubungaku [National language and literature] 14:11,
1-17.
Tokieda, M . 1937b. 'Bun no gainen nitsuite' [On the concept of the sentence]
Part 2. Kokugo to kokubungaku [National language and literature] 14:12,
55-71.
Tokieda, M . 1941. Kokugogakugenron [Principles of national linguistics].
Tokyo: Iwanami shoten.
Tokieda, M . 1950. Nihonbunpoo: koogohen [Japanese grammar: the spoken
language]. Tokyo: Iwanamishoten.
Tokieda, M . 1961. Gendaigo no tokushitsu [Characteristics of the modem
language]. In M . Tokieda (ed.), Kokugo no hataraki to miryoku [The
function and appeal of the national language ] Gendai no hanashikata to
bunshoo 1 [Modem speaking and writing 1]. Tokyo: Koodansha. 15-28.
Tokieda, M . 1968. 'Tokiedabunpoo' no seiritsu to sono genryuu: Suzuki
Akira to dentooteki gengokan [The origin and establishment of 'Tokieda
grammar': Suzuki Akira and the traditional view of language]. In A.
Matsumura et al. (eds), Bunpooron no tenkai [The development of
grammatical theory] Kooza Nihongo no Bunpoo 1] [A Course in Japanese
Grammar 1]. Tokyo: Meijishoin. 1-28.
Wundt, W. 1904 [1902]. Principles of physiological psychology [Trans. E . B.
Titchener]. London: Swan Sonnenschein.
Yamada, Y . 1908. Nihonbunpooron [Theory of Japanese grammar]. Tokyo:
Hoobunkan.
Yamada, Y . 1931. Nihonbunpooyooron [Essential theory of Japanese
grammar]. Tokyo: Iwanamishoten.
Yamada, Y . 1936. Nihonbunpoogakugairon [An outline of Japanese
grammatical theory]. Tokyo: Hoobunkan.
Yamada, Y . 1942. Nihonbunpookoogi [Lectures on Japanese grammar] 4th
ed. Tokyo: Hoobunkan.