In our discussion of the English perfect we noted that some languages have a perfect marker which expresses only the existential/experiential sense. Mandarin is one such language. The meaning of the verbal suffix -guo is in some ways the polar opposite of the meaning of the perfect marker in Baraïn. While the perfect in Baraïn can express all of the standard perfect readings except the experiential, -guo expresses only the experiential perfect. While the perfect in Baraïn requires that the result state of the event still holds true at the time of speaking, -guo requires that the result state no longer holds true at the time of speaking.
The meaning of Mandarin -guo is similar in many ways to the existential/experiential perfect in English; but there are important differences as well. Chao (1968) refers to the suffix -guo as a marker of “indefinite past aspect”. Li & Thompson (1981: 226) identify -guo as a marker of “experiential aspect”, stating that it indicates that the situation has been experienced at least once, at some indefinite time in the past.10 They provide the following minimal pair illustrating the contrast between the perfective (32a), in which the described event occurs within Topic Time, vs. the experiential (32b), in which the described event occurs at some arbitrary time prior to Topic Time.11
(32) a. Nǐ kànjian-le wǒ=de yǎnjìng ma?
2sg see-pfv 1sg=poss glasses Q
‘Did you see my glasses?’ (recently; I’m looking for them)
b. Nǐ kànjian-guo wǒ=de yǎnjìng ma?
2sg see-exper 1sg=poss glasses Q
‘Have you ever seen my glasses?’
Wu (2009) states: “an eventuality presented by -guo is temporally independent of others in the same discourse.” This constraint follows from the fact that normally -guo has indefinite time reference, and so does not establish a new Topic Time to which other clauses or sentences can refer. As a result, clauses marked with -guo are not interpreted as a narrative sequence of events. Iljic (1990: 308) provides the following contrast between the two verbal suffixes –le and -guo, showing that a series of clauses marked with –le is interpreted as a chronological sequence, while the same series of clauses marked with -guo is interpreted as a mere inventory of activities.
(33) a. Qùnián wǒ zuò-le mǎimài, xué-le jìsuànjī, shàng-le yèdàxué.
last.year 1sg do-pfv business study-pfv computer go-pfv evening.university
‘Last year I did some business, (then) studied computers, (then) attended evening university.’ (chronological perspective)
b. Qùnián wǒ zuò-guo mǎimài, xué-guo jìsuànjī, shàng-guo yèdàxué.
last.year 1sg do-exper business study-exper computer go-exper evening.university
‘Last year I did some business, (and) studied computers, (and) attended evening university.’ (inventory perspective)
Examples like (34) are sometimes cited as counter-examples to the generalization that -guo marks indefinite time in the past. The speaker in this sentence is clearly not just claiming to have eaten food at some time in the past, but rather is stating that he has finished eating the most recently scheduled meal.
(34) Wǒ chī-guò fàn le.
1sg eat-finish rice cos12
‘I have already eaten.’ (Ma 1977)
Chao (1968: 251), Comrie (1976: 59) and Xiao & McEnery (2004: 139ff.) state that the -guò in such examples is not the aspectual suffix but a verb root occurring as the second member of a compound verb. Both of these forms are derived from the verb guò ‘to pass by’, and both are written with the same Chinese character. However, the aspectual suffix can be distinguished from the compound verb by phonological and morphological evidence. Phonologically, the aspectual suffix is always toneless (i.e., takes neutral tone) whereas the compound verb takes an optional 4th tone, as marked in (34).13 Morphologically, the compound verb -guò can be followed by the perfective suffix -le, whereas the aspectual suffix -guo cannot. Chu (1998: 39–40) shows that temporal adverbial clauses like the first clause of (35) are another context where the compound verb -guò rather than the aspectual suffix -guo is used. Some authors introduce unnecessary complexity into the discussion of aspectual -guo by failing to make this distinction.
(35) Nǐ míngtian kàn-guò jiù zhīdao le.
2sg tomorrow see-finish then know cos
After you see it tomorrow, you will know.’ (G.-t. Chen 1979)
Many authors have noted an interesting semantic restriction on the use of the aspectual suffix -guo: as first observed by Chao (1968:439; cf. Yeh 1996), there must be a “discontinuity” between Situation Time and Topic Time. If the described event produces a result state, the result state must be over before Topic Time, as seen in (36a). We might represent this discontinuity as follows: TSit ∩ TT = ⌀ (here we assume that the result state is included in TSit). Some authors (e.g. Iljic 1990, Yeh 1996) have suggested that this discontinuity effect is merely an “inference”; but examples (37a) and (38a) seem to indicate that the requirement is an entailment and not just an implicature.14
(36) a. Wǒ shuāi-duàn-guo tuǐ.
1sg fall-break-exper leg
‘I have broken my leg (before).’ (It has healed since.)
b. Wǒ shuāi-duàn-le tuǐ.
1sg fall-break-pfv leg
‘I broke my leg.’ (It may be still in a cast.)
(37) a. Tā qùnián dào Zhōngguó qù-guo, (#xiànzai hái zài nàr ne).
3sg last.year to China go-exper now still at here prtcl
‘He has been to China sometime last year (#and is still there now).’
b. Tā qùnián dào Zhōngguó qù-le, (xiànzai hái zài nàr ne).
3sg last.year to China go-pfv now still at there prtcl
‘He went to China last year (and is still there now).’
(38) a. Tā ài-guo Huáng Xiǎojie, (#xiànzai hái ài-zhe tā ne).
3sg love-exper Huang Miss now still love-cont there prtcl
‘He once loved Miss Huang (#and he still loves her now).’
b. Tā ài Huáng Xiǎojie le, (xiànzai hái ài-zhe tā ne).
3sg love Huang Miss cos now still love-contthere prtcl
‘He has fallen in love with Miss Huang (and he still loves her now).’
Interestingly, this discontinuity requirement is (partially?) dependent on the definiteness of the affected argument.15 When the patient is definite, as in (39a), the use of -guo indicates that the result state no longer obtains; but when the patient is indefinite, as in (39b), there is no such implication/entailment.16
(39) a. Lǐsì nòng-huài-guo zhè bù bǐjìxíng-diànnǎo.
Lisi make-broken-exper this CL laptop
‘Lisi has broken this laptop before.’
(strongly implies that the laptop has been fixed at the time of speech)
b. Lǐsì nòng-huài-guo yī bù bǐjìxíng-diànnǎo.
Lisi make-broken-exper one CL laptop
‘Lisi has broken a laptop before.’
(no commitment as to whether the laptop has been fixed or not)
A number of authors17 have claimed that the situation marked by -guo must be repeatable. If it is an event, there must be a possibility for the same kind of event to happen again. This is a well-known property of the experiential (or existential) perfect in English, and its applicability to -guo is supported by examples like (40), from Ma (1977: 15). However, this claim has been challenged by number of other authors.18 Consider the contrast in (41). Neither being old nor young are states that are repeatable for a single individual. The contrast between the two sentences seems best explained in terms of the discontinuity requirement: a person who is no longer young can still be alive, but not a person who is no longer old.
(40) * Tā sǐ-guo.
3sg die-exper
(intended: ‘He has died before.’)
(41) a. Nǐ yě niánqīng-guo.
you also young-exper
‘You also have been young before.’
b. *Nǐ yě lǎo-guo.
you also old-exper
‘You have also been old before.’
It appears that all of the data which has been proposed in support of the repeatability hypothesis can equally well be explained in terms of the discontinuity requirement. Support for the idea that discontinuity, rather than repeatability, is the operative factor comes from the observation that “repeatability” effects are sensitive to definiteness in exactly the same way as demonstrated above for the discontinuity requirement; this is illustrated in (42).19 The fact that it is possible to use -guo when talking about the actions of dead people, as in (42b), gives further support to the claim that there is no repeatability requirement in Mandarin. Such examples are normally quite unnatural in the English experiential/ existential perfect.
(42) a. *Columbus fāxiàn-guo měizhōu.
Columbus discover-exper America
(intended: ‘Columbus has discovered America before.’)
b. Columbus fāxiàn-guo yī gè xiǎo dǎo.
Columbus discover-exper one CL small island
‘Columbus has discovered a small island before.’
It is useful to compare the semantic effect of the aspectual suffix -guo in various situation types (Aktionsart). With stative predicates, -guo indicates that the state no longer exists (43).20 Therefore, permanent states cannot normally be marked with -guo (44).21
(43) a. Zhāng Xiǎojie guòqù pàng-guo.
Zhang Miss in.past fat-exper
‘Miss Zhang has been fat.’ (implying she is not fat now)
b. Měiguo níuròu yě guì-guo.
America beef also expensive-exper
‘Beef America has also been expensive (in the past but not now).’
c. Tā zài Zhōngguó zhù-guo sān nián.
3sg at China live-exper three year
‘He has lived in China for three years before (but does not live there now).’
(44) *Dāngdì nóngmín zhīdào-guo nà gèzhā yǒudú.
local farmer know-exper that chrome.dreg poisonous
(intended: ‘Local farmers knew that those chrome dregs were poisonous.’)
With atelic events such as activities (45) and non-culminating accomplishments (46), the suffix -guo has the same interpretation as the perfective suffix –le, indicating that the event has terminated.22
(45) Lǐsì dǎ-guo wǎngqiú.
Lisi play-exper tennis
‘Lisi has played tennis before.’
(46) a. #Wǒ xiě-guo gěi Wáng de xìn, hái zài xiě.
1sg write-guo to wang lnk letter, still prog write
‘I wrote Wang’s letter and am still writing it.’
b. Wǒ xiě-guo gěi Wáng de xìn, kěshì méi xiě-wán.
1sg write-guo to wang lnk letter, but not write-finish
‘I wrote Wang’s letter but didn’t finish it.’
In light of what we have said above, we would predict that the aspectual suffix -guo cannot occur with telic predicates whose result state is permanent, because this would mean that discontinuity with Topic time is impossible. This prediction turns out to be true when the patient (or affected argument) is definite. However, as noted above, the discontinuity requirement does not apply when the patient is indefinite; so the aspectual suffix -guo is possible in such contexts.
The examples in (47a–b) contain a Result Compound Verb (RCV), which means that the culmination of the event is entailed. As predicted, -guo is not allowed when the object NP is definite (47a), but is possible when the object NP is indefinite (47b). However, (47c) contains the simple root ‘kill’ with no RCV, and so the culmination of the event would normally be implicated but not entailed. In this example, -guo functions as an explicit indicator that the result state was not achieved.23
(47) a. * Tā shā-sǐ-guo nèi-ge rén.
3sg kill-die-exper that-class person
(intended: ‘He has killed that person.’)
b. Tā shā-sǐ-guo sān-ge rén.
3sg kill-die-exper three-class person
‘He has killed three people.’
c. Tā shā-guo nèi-ge rén.
3sg kill-exper that-class person
‘He tried (at least once) to kill that person (without success).’
A similar pattern is seen in (48). The aspectual suffix -guo can occur with the predicate ‘die’ only when the patient is indefinite (48c). In (48d), which Chu (1998) and Xiao & McEnery (2004) describe as a figurative use of the word ‘die’, -guo functions as an indicator that the result state was not achieved.
(48) a. * Tā sǐ-guo.
3sg die-exper
(intended: ‘He has died before.’) (Ma 1977: 15)
b. Tā sǐ-le.
3sg die-pfv
‘He died.’ (Ma 1977: 15)
c. Yǒu rén zài zhè tiáo hé lǐ yān-sǐ-guo.
have person at this CL river in drown-die-exper
‘Someone has drowned in this river (before).’ (Yeh 1996: 163)
d. Wǒ sǐ-guo hǎojǐ cì.
1sg die-exper quite.a.few time
‘I almost died quite a few times.’ (Chu 1998: 41)
Huang & Davis (1989: 151) point out that -guo can also be used to indicate partial affectedness of a definite object, another way in which the culmination of the event might not be achieved:
(49) a. Gǒu gāngcái chī-le nǐ de píngguǒ.
dog just.now eat-pfv you poss apple
‘The dog just ate your apple.’
b. Gǒu gāngcái chī-guo nǐ de píngguǒ.
dog just.now eat-exper you poss apple
‘The dog just took a bite of your apple.’
10 Some authors take the term “experiential aspect” quite literally, assuming that an animate experiencer must be involved. For example, Xiao & McEnery (2004: 144) write: “The distinguishing feature of -guo is that it conveys a mentally experienced situation.” C. Smith (1997: 267) states that “sentences with -guo ascribe to an experiencer the property of having participated in the situation.” However, -guo can also be used in clauses which contain no animate arguments.
11 Examples from Ma (1977: 19); Li & Thompson (1981: 227).
12 The abbreviation cos stands for ‘change-of-state’, the label used by Soh (2009) for the sentencefinal particle which indicates that a situation is currently true but was not true in the past. Li & Thompson (1981: 238ff.) use the label ”Currently Relevant State” for this particle.
13 Comrie states that this 4th tone is optional but is usually pronounced.
14 Examples (36)–(38) come from Ma (1977: 18, 25) and Chao (1968).
15 Lin (2007); Wu (2008); C.-c. Chen (2009).
16 Examples from C.-c. Chen (2009); cf. Lin (2007).
17 Ma (1977); Li & Thompson (1981: 230); Yeh (1996); C. Smith (1997: 268).
18 G.-t. Chen (1979), Iljic (1990), Xiao & McEnery (2004: 147–148), Pan & Lee (2004), Lin (2007).
19 Examples from Yeh (1996: 153, 163)
20 Examples from Ma (1977: 20, 23)
21 Example (44) comes from Xiao & McEnery (2004: 149).
22 Examples from C. Smith (1997: 267).
23 The examples in (47) come from Ma (1977: 23).
