Page 1 of 11
Advances in Social Sciences Research Journal – Vol. 8, No. 11
Publication Date: November 25, 2021
DOI:10.14738/assrj.811.11275. Kang, N. (2021). Differences Between Korean Null Arguments and Korean Overt Pronouns. Advances in Social Sciences Research
Journal, 8(11). 378-388.
Services for Science and Education – United Kingdom
Differences Between Korean Null Arguments and Korean Overt
Pronouns
Namkil Kang
Far East University, South Korea
ABSTRACT
The ultimate goal is this paper is to provide five pieces of evidence that Korean null
pronouns are not semantically and syntactically equivalent to Korean overt
pronouns. First, when Korean overt pronouns and Korean null pronouns have the
only NP as their antecedent, the truth condition becomes different. Second, when
Korean overt pronouns and Korean null pronouns take the even NP as their
antecedent, the truth condition becomes different. Third, the Korean overt pronoun
ku ‘he’ is associated with its antecedent by coreference, whereas Korean null
pronouns are associated with their antecedent by binding. Fourth, Korean overt
pronouns yield a strict reading, whereas Korean null arguments induces the
strict/sloppy ambiguity. This in turn suggests that Korean null pronouns yield
looser interpretations than Korean overt pronouns. Additionally, it is worth noting
that Korean overt pronouns induce a definite reading, whereas Korean null
pronouns yield indefinite and definite readings. Fifth, Korean overt pronouns and
Korean null pronouns are not alike in that the former is sensitive to phi-features
(gender, number, and person), whereas the latter is not. This paper argues that
Korean null pronouns are not syntactically the equivalent of Korean overt
pronouns.
Keywords: null pronoun, overt pronoun, null argument, binding, coreference, sloppy
identity, definite reading
INTRODUCTION
The main goal of this paper is to provide evidence that Korean null pronouns are not
semantically and syntactically equivalent to Korean overt pronouns. The organization of this
paper is as follows. In section 2.1, we show that when Korean overt pronouns and Korean null
pronouns have the only NP as their antecedent, the truth condition becomes different. In
section 2.2, we argue that when Korean overt pronouns and Korean null pronouns take the even
NP as their antecedent, the truth condition becomes different. In section 2.3, we further argue
that the Korean overt pronoun ku ‘he’ is associated with its antecedent by coreference, whereas
Korean null pronouns are associated with their antecedent by binding. In section 2.4, we
maintain that Korean overt pronouns yield a strict reading, whereas Korean null arguments
induces the strict/sloppy ambiguity. This in turn suggests that Korean null pronouns yield
looser interpretations than Korean overt pronouns. Additionally, we show that Korean overt
pronouns induce a definite reading, whereas Korean null pronouns yield indefinite and definite
readings. In section 2.5, we argue that Korean overt pronouns and Korean null pronouns are
not alike in that the former is sensitive to phi-features (gender, number, and person), whereas
the latter is not. We demonstrate that Korean null arguments are not syntactically equivalent
to Korean overt pronouns.
Page 2 of 11
379
Kang, N. (2021). Differences Between Korean Null Arguments and Korean Overt Pronouns. Advances in Social Sciences Research Journal, 8(11).
378-388.
URL: http://dx.doi.org/10.14738/assrj.811.11275
DIFFERENCES BETWEEN KOREAN OVERT PRONOUNS AND KOREAN NULL ARGUMENTS
The Only NP and the Truth Condition
The main goal of this section is to verify that in some environments, Korean overt pronouns and
Korean null pronouns are synonymously used, but the former is not equivalent to the latter. In
this section, we provide evidence that Korean null pronouns are not semantically equivalent to
Korean overt pronouns. We show that when Korean overt pronouns and Korean null pronouns
have the only NP as their antecedent, the truth condition becomes different.
Now let us consider (1) and (2):
(1) Johni-mani kui-ka ikilkessila-ko sayngkakhanta.
only he-NOM will win-COMP think
(Only John thinks that he will win.)
(2) Johni-mani ei ikilkessila-ko sayngkakhanta.
only will win-COMP think
(Only John thinks that he will win.)
When Korean overt pronouns and Korean null pronouns have the only NP as their antecedent,
the truth condition becomes different. In some environments, Korean overt pronouns and
Korean null pronouns are interchangeable, but the following state of affairs suggests the
opposite:
(3) =0
John John
Jim Jim
Bill Bill
Tom Tom
In (3), John is not the only person who thinks that he will win. In (3), Jim, Bill, and Tom think
that they themselves will win. This is not compatible with (1). In (3), John is not the only one
who thinks that he will win, which renders (1) false.
Now let us consider the following state of affairs:
(4) =0
John John
Mary Mary
Bill Bill
Tom Tom
We understand (1) as false since in (4), John is not the only person who thinks that he will win.
That is to say, Mary, Bill, and Tom think that John will win, which is not compatible with (1).
Now let us observe (5):
(5) =1
John John
Mary Mary
Bill Bill
Tom Tom
Page 3 of 11
380
Advances in Social Sciences Research Journal (ASSRJ) Vol. 8, Issue 11, November-2021
Services for Science and Education – United Kingdom
In (5), John is the only person who thinks that he will win. Mary, Bill, and Tom do not think that
they themselves will win. In addition, they do not think that John or someone else will win.
Thus, (5) is compatible with (1), which renders (1) true.
Now let us observe (2) and (6):
(6) =0
John John
Mary Mary
Bill Bill
Tom Tom
In (6), John is not the only person who is likely to win. That is to say, Mary, Bill, and Tom think
that they themselves will win. We take (2) as indicating that John is the only person who is likely
to win. This is not compatible with the state of affairs in (6), which renders (2) false.
Now observe the following state of affairs:
(7) =0
John John
Mary Mary
Bill Bill
Tom Tom
In (7), John is not the only person who is likely to win. That is to say, Mary thinks that Bill will
win and Bill thinks that Mary will win, which is not compatible with (2). The meaning of (2) is
‘John is the only person who is likely to win’.
Now let us consider the following state of affairs:
(8) =1
John John
Mary Mary
Bill Bill
Tom Tom
The state of affairs in (8) is in accordance with the meaning of (2). More specifically, in (8), John
is the only person who is likely to win. John thinks that he will win and Mary, Bill, and Tom think
that John will win. Thus, the state of affairs in (8) renders (2) true. This in turn suggests that
Korean null pronouns are not semantically equivalent to Korean overt pronouns and that
Korean null pronouns are null arguments different from overt pronouns.
The Even NP and the Truth Condition
In this section, we show that when Korean overt pronouns and Korean null pronouns take the
even NP as their antecedent, the truth condition becomes different. Let us consider the
following examples:
(9) John cocha ku-ka ttokttokhata-ko sayngkakhanta.
even he-NOM intelligent-COMP think
(Even John thinks that he is intelligent.)
Page 4 of 11
381
Kang, N. (2021). Differences Between Korean Null Arguments and Korean Overt Pronouns. Advances in Social Sciences Research Journal, 8(11).
378-388.
URL: http://dx.doi.org/10.14738/assrj.811.11275
(10) John cocha e ttokttokhata-ko sayngkakhanta.
even intelligent-COMP think
(Even John thinks that he is intelligent.)
Now let us observe the following state of affairs:
(11) =1
John John
Mary Mary
Bill Bill
Tom Tom
The state of affairs in (11) renders (9) true since Mary, Bill, and Tom think that John is
intelligent and even John thinks that he himself is intelligent. The meaning of (9) is ‘everyone
thinks that John is intelligent and John is the last person who thinks that he is intelligent’.
However, let us consider the state of affairs in (12):
(12)=0
John John
Mary Mary
Bill Bill
Tom Tom
The meaning of (12) is ‘for every x, x thinks that x is intelligent’. However, (9) indicates that
everyone thinks that John is intelligent and John is the last person who thinks that he himself is
intelligent. Thus, this is not compatible with the state of affairs in (12).
Now let us consider (10) and the state of affairs in (13):
(13) =1
John John
Mary Mary
Bill Bill
Tom Tom
The meaning of (10) is ‘for every x, x thinks that x is intelligent and John is the last person who
thinks that he is intelligent’. Thus, the state of affairs in (13) renders (10) true.
Now consider the state of affairs in (14):
(14) =0
John John
Mary Mary
Bill Bill
Tom Tom
The state of affairs in (14) indicates that everyone thinks that John is intelligent. However, (10)
clearly indicates that for every x, x thinks that x is intelligent and John is the last person who
thinks that he is intelligent. This is not compatible with the state of affairs in (14). Thus, (14)
Page 5 of 11
382
Advances in Social Sciences Research Journal (ASSRJ) Vol. 8, Issue 11, November-2021
Services for Science and Education – United Kingdom
renders (10) false. This in turn suggests that Korean null pronouns are not equivalent to Korean
overt pronouns.
Binding and Coreference
In what follows, we wish to argue that the Korean overt pronoun ku ‘he’ is associated with its
antecedent by coreference, whereas Korean null subjects are associated with their antecedent
by binding. Reinhart (1983, 1993) argues that Binding Theory regulates the distribution of
bound anaphora, not coreference and that coindexed anaphora allows binding. Let us begin
making a difference between binding and coreference. Let us observe (15):
(15) Tom-cocha ku-casin-ul piphanhayssta.
even he-self-ACC criticized
(Even Tom criticized himself.)
(a) Most men criticized themselves (binding)
(For most men x, x criticized x and Tom is the last person who criticized himself.
(b) Most men criticized Tom (coreference)
(For most men x, x criticized Tom, and Tom is the last person who criticized himself.
As illustrated in (15), the Korean reflexive ku-casin ‘se-self’ is associated with its antecedent
either by binding or by coreference. More specifically, the Korean reflexive ku-casin ‘he-self’
induces a bound variable reading, whereas it can be associated with its antecedent through
coreference. The following sentences also provide a clear distinction between coreference and
binding in the non-local context with Korean overt pronouns and Korean null pronouns:
(16) Tomi-cocha kui-ka kui-uy tongsayng-ul piphanhayssta-ko
even he-NOM he-GEN brother-ACC criticized
malhayssta said
(Even Tom said that he criticized his brother.)
Most men said that they criticized Tom’s brother (coreference)
(For most men x, x said that x criticized Tom’s brother and Tom is the last person who criticized
his brother.)
(17) Tomi-cocha ei kui-uy tongsayng-ul piphanhayssta-ko
even he-GEN brother-ACC criticized
malhayssta said
(Even Tom said that he criticized his brother.)
Most men said that they criticized their brother (binding)
(For most men x, x said that x criticized x’s brother and Tom is the last person who criticized
his brother.)
As indicated in (16) and (17), the Korean overt pronoun ku ‘he’ is associated with its antecedent
by coreference. On the other hand, the Korean null subject is associated with its antecedent by
binding. This in turn suggests that in some environments, Korean overt pronouns and Korean
null pronouns are synonymously used, but the former is not equivalent to the latter.
The same applies to QP/WH-antecedents. The assumption that Korean overt pronouns and
Korean null pronouns are non-locally bound is widely accepted in the literature:
(18) Tomi-i kui-ka ttokttokhata-ko sayngkakhanta.
NOM he-NOM intelligent-COMP think
Page 6 of 11
383
Kang, N. (2021). Differences Between Korean Null Arguments and Korean Overt Pronouns. Advances in Social Sciences Research Journal, 8(11).
378-388.
URL: http://dx.doi.org/10.14738/assrj.811.11275
(Tom thinks that he is intelligent.)
(19) Tomi-i ei ttokttokhata-ko sayngkakhanta.
NOM intelligent-COMP think
(Tom thinks that he is intelligent.)
When a DP occurs as the antecedent, there is no difference in naturalness between Korean overt
pronouns and Korean null pronouns. However, when a QP appears as the antecedent, the
difference in naturalness between Korean overt pronouns and Korean null pronouns is sharper:
(20) *Nwukwunai kui-ka ttokttokhata-ko sayngkakhanta.
everyone he-NOM intelligent-COMP think
(Everyone thinks that he is intelligent.)
(21) Nwukwunai ei ttokttokhata-ko sayngkakhanta.
everyone intelligent-COMP think
(Everyone thinks that he is intelligent.)
The ungrammaticality of (20) indicates that binding is not available for the Korean overt
pronoun ku ‘he’, whereas binding is available for the Korean null subject. More specifically, the
Korean null subject gives rise to a bound variable reading, whereas the Korean overt pronoun
ku ‘he’ does not. An important question that naturally arises is “why is (18) grammatical?” We
wish to argue that in (18), the Korean overt pronoun ku ‘he’ is associated with its antecedent
not through binding but through coreference. Thus, it is reasonable to conclude that Korean
overt pronouns are associated with their antecedent by coreference, whereas Korean null
pronouns are associated with their antecedent by binding. This in turn suggests that Korean
null pronouns are not pro but null arguments different from Korean overt pronouns.
Sloppy Identity and a Definite Reading
The so-called sloppy identity has been discussed in Keenan (1971), Sag (1976), Williams
(1977), Partee (1978), Reinhart (1983), Heim and Kratzer (1997), Bϋring (2005), Mckillen
(2016), and others. In (22), the second conjunct is two ways ambiguous between (a) and (b):
(22) Tom admired his teacher and so did Mary.
(a) On the strict reading
Tom admired Tom’s teacher and Mary admired Tom’s teacher.
Tom (λx (x admired his (=Tom’s) teacher))
Mary (λx (x admired his (=Tom’s) teacher))
(b) On the sloppy reading
Tom admired Tom’s teacher and Mary admired Mary’s teacher.
Tom (λx (x admired x’s teacher))
Mary (λx (x admired x’s teacher))
As alluded to in (22), the second conjunct induces sloppy and strict readings under VP ellipsis.
In order to verify that Korean null pronouns are not semantically equivalent to Korean overt
Page 7 of 11
384
Advances in Social Sciences Research Journal (ASSRJ) Vol. 8, Issue 11, November-2021
Services for Science and Education – United Kingdom
pronouns, we are now in a position to consider Korean null pronouns and Korean overt
pronouns under sloppy identity.
(23) a. Tom-un palamkwa hamkkey salacita-lul ilkessta.
TOP wind with gone-ACC read
(Tom read gone with the wind.)
b. Bill-to ku-kess-ul ilkessta.
too it-ACC read
(Bill read it, too.)
c. Mary-to e ilkessta.
too read
(Mary read it, too.)
(23b) means that Bill read the same book (gone with the wind) Tom read (a strict reading).
Likewise, (23c) means that Mary read the same book (gone with the wind) Tom and Bill read
(a strict reading). This seems to suggest that Korean overt pronouns and Korean null pronouns
are synonymously used. However, (24) provides further confirmation that Korean null
pronouns are not semantically equivalent to Korean overt pronouns:
(24) a. Tom-un twukwen-uy chayk-ul ilkessta.
TOP two-GEN book-ACC read
(Tom read two books.)
b. Bill-to ku-kess-tul-ul ilkessta.
too them-ACC read
(Bill read them, too.)
(25) a. Tom-un twukwen-uy chayk-ul ilkessta.
TOP two-GEN book-ACC read
(Tom read two books.)
b. Mary-to e ilkessta.
too read
(Mary read them, too.)
(24b) means that Bill read two books Tom read (a strict reading). On the other hand, (25b)
means that Mary read two books Tom read (a strict reading). In addition, (25b) means that
Mary read two books Tom did not read (a sloppy reading). In a word, (25b) induces the
strict/sloppy ambiguity. This in turn suggests that Korean null pronouns yield looser
interpretations than Korean overt pronouns. Thus, it is reasonable to assume that Korean null
pronouns are not semantically equivalent to Korean overt pronouns.
The fact that Korean overt pronouns induce a definite reading, whereas Korean null pronouns
yield indefinite and definite readings lends its support to the assumption that Korean null
pronouns are not semantically equivalent to Korean overt pronouns:
(26) a. Tom-un caymiissnun yenghwa-lul hanpyen
TOP interesting movie-ACC a
poassta.
saw
(Tom saw an interesting movie.)
b. Bill-to ku-kess-ul poassta.
Page 8 of 11
385
Kang, N. (2021). Differences Between Korean Null Arguments and Korean Overt Pronouns. Advances in Social Sciences Research Journal, 8(11).
378-388.
URL: http://dx.doi.org/10.14738/assrj.811.11275
too it-ACC saw
(Bill saw an interesting movie, too.)
(27) a. Tom-un caymiissnun yenghwa-lul hanpyen
TOP interesting movie-ACC a
poassta.
saw
(Tom saw an interesting movie.)
b. Bill-to e poassta.
too saw
(Bill saw an interesting movie, too.)
In (26), the Korean overt pronoun ku-kess ‘it’ is interpreted to mean that Bill saw an interesting
movie and that Tom and Bill saw the same movie. This is something corresponding to the
English definite pronoun it. On the other hand, the null argument e indicates that Bill saw an
interesting movie that Tom did not see. In this case, the referent of the null argument e is
something corresponding to the English indefinite pronoun one. In addition, the null argument
e indicates that Bill saw the same movie that Tom saw. Thus, the referent of the null argument
e is something to the English definite pronoun it. To sum up, Korean overt pronouns yield a
definite reading, whereas Korean null pronouns give rise to indefinite and definite readings.
Likewise, the following sentences clearly show that Korean overt pronouns yield a definite
reading, whereas Korean null pronouns give rise to indefinite and definite readings:
(28) a. Tom-un sekwen-uy chayk-ulilkessta.
TOP three-copies-GEN book-ACC read
(Tom read three books.)
b. Bill-to ku-kess-tul-ul ilkessta.
too them-ACC read
(Bill read them, too.)
(29) a. Tom-un sekwen-uy chayk-ulilkessta.
TOP three copies-GEN book-ACC read
(Tom read three books.)
b. Bill-to e ilkessta.
too read
(Bill read them, too.)
In (28), the overt pronoun ku-kess-tul ‘they’ indicates that Tom and Bill read the same three
books. This is something corresponding to the English definite pronoun it. In (29), on the other
hand, the null argument e indicates that Bill read three books that Tom did not read. This is
something corresponding to the English indefinite pronoun one. In addition, it is interpreted as
‘Tom and Bill read the same three books’, which is something corresponding to the English
definite pronoun it. We thus conclude that Korean overt pronouns yield a definite reading,
whereas Korean null pronouns give rise to indefinite and definite readings. Simply put, Korean
null pronouns yield looser interpretations than Korean overt pronouns. This in turn implies
that Korean null pronouns are not semantically equivalent to Korean overt pronouns.
Page 9 of 11
386
Advances in Social Sciences Research Journal (ASSRJ) Vol. 8, Issue 11, November-2021
Services for Science and Education – United Kingdom
Phi-features and Agreement
Korean overt pronouns and Korean null pronouns are not alike in that the former is sensitive
to phi-features (gender, number, and person), whereas the latter is not. We demonstrate that
Korean null pronouns are syntactically equivalent to Korean overt arguments. Let us consider
the following sentences:
(30) Nwu-ka ikilkessila-ko ne-nun sayngkakhani?
Who-NOM will win-COMP you-TOP think
(Who do you think will win?)
(31) a. Ku [points to Tom]-ka ikilkeya.
he-NOM will win
(Tom will win.)
b. e [points to Tom] ikilkeya.
will win
(Tom will win.)
As shown in (31a) and (31b), the speaker employs the pointing effect in order to denote a
particular entity or entities. The Korean overt pronoun ku ‘he’ can refer to Tom. Likewise, the
null argument e can also refer to Tom, which depends on the speaker’s intention. This seems to
suggest that Korean overt pronouns and Korean null arguments are synonymously used.
However, the following sentences indicate that Korean null arguments are not syntactically
equivalent to Korean overt arguments:
(32) Nwu-ka ikilkessila-ko ne-nun sayngkakhani?
Who-NOM will win-COMP you-TOP think
(Who do you think will win?)
(33) a. *Ku [points to Tom and Bill]-ka ikilkeya.
he-NOM will win
(Tom and Bill will win.)
b. e [points to Tom and Bill] ikilkeya.
will win
(Tom and Bill will win.)
As indicated in (33a), the Korean overt pronoun ku ‘he’ cannot refer to Tom and Bill since it is a
singular dependent term. On the other hand, the null argument e can refer to Tom and Bill. The
pointing effect does not work for (33a), which in turn indicates that the overt pronoun ku ‘he’
is sensitive to the number feature. On the other hand, the pointing effect works for (33b), which
in turn implies that the null argument e is not sensitive to the number feature. In a word, Korean
null arguments are not exactly the syntactic equivalent of Korean overt pronouns. The same
applies to the gender feature:
(34) a. *Tom-i ku-nye [points to Jim]-lul piphanhaysse.
NOM she-ACC criticized
(Tom criticized Jim.)
b. Tom-i e [points to Jim] piphanhaysse.
NOM criticized
(Tom criticized Jim.)
Page 10 of 11
387
Kang, N. (2021). Differences Between Korean Null Arguments and Korean Overt Pronouns. Advances in Social Sciences Research Journal, 8(11).
378-388.
URL: http://dx.doi.org/10.14738/assrj.811.11275
As alluded to in (34a), the Korean overt pronoun ku-nye ‘she’ cannot refer to Jim. The
ungrammaticality of (34a) is attributed to the fact that ku-nye ‘she’ does not agree in gender
with Jim. That is, Korean overt pronouns are sensitive to the gender feature. However, we
attribute the grammaticality of (34b) to the fact that Korean null arguments are not sensitive
to the gender feature. This in turn suggests that Korean null arguments have more functional
load than Korean overt pronouns.
Finally, let us consider the following sentences:
(35) a. *Tom-i ku [points to us]-lul piphanhaysse.
NOM he-ACC criticized
(Tom criticized us.)
b. Tom-i e [points to us] piphanhaysse.
NOM criticized
(Tom criticized us.)
We attribute the ungrammaticality of (35a) to person disagreement between ku ‘he’ and us,
which indicates that Korean overt pronouns are sensitive to the person feature. On the other
hand, the null argument e can refer to us, which suggests that Korean null arguments are not
sensitive to the person feature. From all of this, it is clear that Korean overt pronouns are
sensitive to phi-features, whereas Korean null arguments are not. We thus conclude that
Korean null arguments have more functional load than Korean overt pronouns.
CONCLUSION
The main purpose of this paper is to provide evidence that Korean null pronouns are not
semantically and syntactically equivalent to Korean overt pronouns. In section 2.1, we have
shown that when Korean overt pronouns and Korean null pronouns have the only NP as their
antecedent, the truth condition becomes different. In section 2.2, we have argued that when
Korean overt pronouns and Korean null pronouns take the even NP as their antecedent, the
truth condition becomes different. In section 2.3, we have maintained that the Korean overt
pronoun ku ‘he’ is associated with its antecedent by coreference, whereas Korean null pronouns
are associated with their antecedent by binding. Reinhart (1983, 1993) argues that Binding
Theory regulates the distribution of bound anaphora, not coreference and that coindexed
anaphora allows binding. In section 2.4, we have contended that Korean overt pronouns yield
a strict reading, whereas Korean null arguments induces the strict/sloppy ambiguity. This in
turn suggests that Korean null pronouns yield looser interpretations than Korean overt
pronouns. In addition, we have shown that Korean overt pronouns induce a definite reading,
whereas Korean null pronouns yield indefinite and definite readings. In section 2.5, we have
argued that Korean overt pronouns and Korean null pronouns are not alike in that the former
is sensitive to phi-features (gender, number, and person), whereas the latter is not. We have
demonstrated that Korean null arguments are not syntactically equivalent to Korean overt
pronouns.
Page 11 of 11
388
Advances in Social Sciences Research Journal (ASSRJ) Vol. 8, Issue 11, November-2021
Services for Science and Education – United Kingdom
References
[1]. Bϋring, D. (2005). Binding Theory. Cambridge University Press.
[2]. Heim, I and Kratzer, A (1998). Semantics in Generative Grammar. Blackwell Publishers.
[3]. Keenan, E. (1971). Names, Quantifiers and a Solution to the Sloppy Identity Problem. Papers in Linguistics 4 (2):
159-178.
[4]. McKillen, A. (2016). VP Ellipsis and Sloppy Readings. ms. McGill University.
[5]. Partee, B. H. (1978). Bound Variables and Other Anaphors. In D. Waltz, ed., Proceedings of TINLAP. University of
Illinois.
[6]. Reinhart, T. (1983). Anaphora and Semantic Interpretation. Croom Helm, London.
[7]. Reinhart, T and Reuland, E. (1993). Reflexivity. Linguistic Inquiry 24: 657-720.
[8]. Sag, I. (1976). Deletion and Logical Form. Doctoral Dissertation, MIT.
[9]. Williams, E. (1977). Discourse and Logical Form. Linguistic Inquiry 8 (1): 445-479.