• Ingen resultater fundet

97

98

system of disjunctive particles within the different languages, nor are they intended to. Instead they are included to add a bit of perspective to the discussion of the understanding of ‘or’ in this context. I understand that this will naturally have an impact on the type of conclusions I will be able to draw from my discussion.

Furthermore, I want to add a quick note about the terminology I use for this discussion. Within logic, i.e. formal semantics, there is a distinction between conjunction (˄) on the one hand, in English represented by the conjunctive particle ‘and’, and disjunction (˅) on the other, represented in English by the disjunctive particle ‘or’. Many grammars and dictionaries (e.g. both the Oxford English Dictionary and the Cambridge Grammar of the English Language), however, treat both

‘and’ and ‘or’ as conjunctions or coordinators as opposed to subordinators (Huddleston, Payne &

Petersen, 2002). I shall use disjunction to refer to ‘or’ and conjunction to refer to ‘and’, keeping with the terms from logic.

The native speakers of English:

Starting with the native speakers of English their answers illustrated in figure 6 below shows an example of what we may call a grammaticalised distinction in native English.

Figure 6 Overview of the answers from the native speakers of English. Numbers of participants selecting a given answer are presented as percentages for ease of comparison across groups, as the different groups have slightly different numbers of participants.

All of the native speakers of British English selected that two people were killed. This clearly shows that the understanding of the disjunctive particle ‘or’ is not an individual choice made on

0%

76%

100%

24%

0%

20%

40%

60%

80%

100%

How many people were killed? How many killers were there?

Proportion of participants

Native speakers of English - overview of both questions

1 person 2 persons

99

background knowledge or context, but rather the choice is governed by pure convention, it is a grammatical distinction. The second question about the number of killers does not show the same kind of agreement. This is perhaps not surprising as the text does not give this information.

Though a majority (76 %) of the participants agree on only one killer, about a fourth of them (24

%) selected two killers. This shows that the distinction here is not governed by a rule – the common code – but is an individual interpretation based on the situational context, including the participant’s previous experiences etc.

But what is really the function of ‘or’ here? The OED describes it as: “conjunction, 1) Used to coordinate two (or more) sentence elements between which there is an alternative. Things so coordinated may differ in nature, or quality, or merely in quantity in which case the one may include the other, as in ‘it will cost a pound or one pound fifty’, ‘two or three minutes’, ‘a word or two’. The second member may also express a correction or modification of the first, which may be strengthened by expanding or to or even, or rather, or at least” (Or, n.d.). The same distinction is found in the Cambridge Grammar: “The relation between and and or is comparable to that between all and some – or universal and existential quantification […] With and we are concerned with a set in totality, whereas with or the members of the set are regarded as alternatives” (Huddleston, Payne & Petersen, 2002, p.1293). I other words, ‘or’ coordinates between alternatives. But this definition makes little sense when looking at how the native speakers understood ‘or’. They did not understand Mrs Hewitt and Dr Gott as alternatives (one or the other) but rather as a whole (one and the other). So how do we explain that ‘or’ was understood as ‘and’?

Within logic, there is a distinction between exclusive disjunction and inclusive disjunction defined according to their truth value (Huddleston, Payne & Petersen, 2002). A proposition with exclusive disjunction is true only if one of the components is false, i.e. for the ‘or’ to be an exclusive disjunction in ‘I have apples or pears’ I would only have apples or only have pears not both (or neither). An inclusive disjunction, on the other hand, is true as long as just one of the components is true, i.e. for the ‘or’ in ‘I have apples or pears’ to be an inclusive disjunction, I would have either apples or pears or both apples and pears (Huddleston, Payne & Petersen, 2002). Applying this logic would suggest that the native speakers of English understand ‘Mrs Hewitt or Dr Gott’

as an inclusive disjunction, not only one is true but both can be, in this case both are in fact, true.

100

While the application of logic sheds some light on the understanding of ‘or’, I am still wondering why they would interpret it as inclusive and what it means that it is inclusive in this case as opposed to if it had just been standard conjunction using ‘and’. Huddleston, Payne & Petersen (2002, p.1297) note that sometimes sentences with ‘or’ will be: “pragmatically equivalent to sentences with and instead of or [and that] the crucial feature is that although they present a choice it doesn’t matter to the speaker which alternative is chosen” (Huddleston, Payne &

Petersen, 2002, p.1297). This would then mean that the participants understand the ‘or’ to be ‘and’

because they pragmatically infer from the surrounding sentences that since the reporter is asking the policeman about who killed Mrs Hewitt or Dr Gott, they must both have been killed, and the reporter just wants information about the killing of any one of them, it does not matter which, hence the use of ‘or’.

There might be another possible explanation as to how the native speakers of English were able to understand ‘or’ as ‘and’. If we return to Lorentzen’s work on disjunction in Danish for a moment, we might find some inspiration in applying Durst-Andersen’s way of thinking. Danish and English are both hearer-oriented languages, which means, quite literally, that they are directed towards the hearer (Durst-Andersen, 2011a). As mentioned in Chapter 4, the utterance (with its different grammatical elements) serves as information to the hearer; it is a signal to the hearer to recreate the situation behind the speaker’s experience of situation. This is how the communication process works for hearer-oriented languages, such as Danish and English. Lorentzen (2016, p.79) highlights that the use of the Danish ‘eller’ (equivalent to ‘or’) is a signal to the hearer that the speaker’s experience of the situations behind are stored in different places in the speaker’s mental storages and the speaker cannot precisely mark where, hence the use of ‘or’. This seems similar to the following example by Huddleston, Payne & Petersen (2002, p.1294): “There is a copy in the office or in the library, for example, is perfectly consistent with both component propositions being true – and indeed I might say it knowing that both are true, using or rather than and because I am thinking of a choice as to which copy to consult”. In other words, ‘or’ in its inclusive use is a signal to the hearer. It carries the information that the speaker’s experience of the situations behind two elements joined by the disjunction are either stored in different places in the speaker’s mental stores or simply have no real match in the speaker’s mental stores.

If we apply this to the case of ‘Reporter: Do you have any idea who might have killed Mrs Hewitt or Dr Gott? Detective: No, unfortunately we have no leads’, then the ‘or’ does not mean that the

101

reporter does not know which of them were killed, it just means that he/she has no specific picture stored about the killing of them, if it was by the same killer or different ones, same place or different places, same time or different times, etc. The native speakers of English know this. They understand the signal conveyed by ‘or’ and the information it carries with it, i.e. that both Mrs Hewitt and Dr Gott were killed but the reporter just does not have any specific picture stored about the details of it in his/her mental stores.

The Japanese speakers of English:

Looking at the Japanese speakers of English, we see from the figure 7 below that the majority have the same comprehension of the text as the native speakers of English in terms of how many people were killed.

Figure 7 Overview of the answers from the Japanese speakers of English. Numbers of participants selecting a given answer are presented as percentages for ease of comparison across groups, as the different groups have slightly different numbers of participants.

However, it does deserve a mention that almost a third of the Japanese participants deviate from this and find that only one person was killed. When I look at the participants who selected only one person killed, nothing really sets them apart from the rest of the Japanese participants. Two are male (out of a total of five), all have English as their only foreign language, and three of the six have had longer stays abroad in English speaking countries (the total amount of Japanese participants who have had stays abroad is nine). Their self-assessed English skills within reading vary from 2 to 6, with the average being 3,9, which is exactly the same as for the entire group.

Since I have not carried out a similar test in Japanese, I have no way of knowing whether the 29

29%

81%

71%

19%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

How many people were killed? How many killers were there?

Proportion of participants

Japanese speakers of English - overview of both questions

1 person 2 persons

102

% is a result of cross-linguistic influence, but it does show that there is not a single, uniform Japanese understanding of the English conjunctions, but rather different understandings.

Some studies have been conducted on the possible transfer of the Japanese L1 understanding of disjunction into English L2, however mainly in relation to negation. An experimental study by Grüter, Lieberman & Gualmini considered possible transfer of the L1 understanding of negation and disjunction into L2 by Japanese learners of English and English learners of Japanese (Grüter, Lieberman, & Gualmini, 2010). The researchers found what they refer to as ‘full transfer’ from Japanese into English by Japanese speakers of English of a medium to high self-assessed proficiency, some even with long stays in English-speaking countries. However, they tested disjunction in relation to negation only. What they show is that a sentence like ‘Mary does not eat broccoli or spinach’ in native English would mean that Mary does not eat broccoli and does not eat spinach (the authors refer to this as the narrow scope of the disjunctive article, i.e. the negation having scope over disjunction, and thus negated disjunction becomes a conjunction condition), but a similar sentence in Japanese with the Japanese disjunction ‘ka’ would mean that she may eat the one but not the other (the authors refer to this as the wide scope of the disjunctive, i.e.

disjunction takes scope over the negation, and this is considered a disjunctive condition).

What their experiments show is that most of the Japanese learners of English transferred their L1 understanding of ‘ka’ into the English ‘or’ in the sense that they accepted a disjunctive condition for sentences where the native speakers of English serving as control group did not Japanese (Grüter, Lieberman, & Gualmini, 2010). Interestingly, the same was not the case for the English learners of Japanese, i.e. most of them accepted a disjunctive condition in the same way the Japanese L1 speakers did Japanese (Grüter, Lieberman, & Gualmini, 2010). The authors conclude that this shows that it is harder to learn the narrow scope of disjunction in relation to negation because this involves unlearning the wide scope: “Not surprisingly, our results show that as a group, English learners of Japanese succeed, while Japanese learners of English fail. This should remind the reader of the substantial body of evidence suggesting that acquiring any given property of the target language is easier than unlearning any property that is transferred from the L1”

(Grüter, Lieberman, & Gualmini, 2010, p.145).

It is important to notice that the experiment above relates to disjunction in relation to negation, i.e. ‘not … or’, which is not directly the case with the text we use. Furthermore, my text differs

103

syntactically from the texts used in the experiments by Grüter, Lieberman & Gualmini. These tested the scope of negation and disjunction in two main clauses with the disjunctive particle appearing in the direct object (‘the horse ate the cake, but he didn’t eat the carrot or the pepper’).

For our text (‘Do you have any idea who killed Mrs Hewitt or Dr Gott) the disjunctive particle does not appear in connection with a negation. Furthermore, for our text the disjunctive particle is in a subordinate clause to ‘any’. Similar to what Grüter, Lieberman & Gualmini (2010) found, Goro & Akiba’s (2004) investigation of the acquisition of disjunction in Japanese argue that the Japanese disjunction ‘ka’ is interpreted with wide scope in simple negative sentences, but in sentential complements or relative clauses it is in fact interpreted as similar to the interpretation of ‘or’. If this is the case, it would mean that the ‘or’ in my text would have an inclusive interpretation also in Japanese. Which is indeed the case for the majority, 71%, but it does not explain why the 29% made an exclusive interpretation. It suggests that for at least some of the Japanese speakers of English, the information carried by ‘or’ signalling that both Mrs Hewitt and Dr Gott were killed but that the reporter just does not have any specific picture stored about the details of it in his/her mental stores is not picked up by these participants. Their comprehension is different, perhaps due to differences in the communication process of their mother tongue.

In relation to the second question regarding the number of killers, the Japanese speakers of English resemble the rest of the groups, including the native speakers, with the vast majority (81%) agreeing that there is only one killer. As mentioned previously, the text does not necessarily give any information as to how many killers there were, which makes it all the more interesting that the Japanese speakers of English seem to agree more on this matter than on the number of victims, which – if they fully understand the English common code – should be given by the text.

Summing up, although the majority of the Japanese speakers of English show comprehension similar to that of the native speakers of English, almost a third do not, but instead seem to understand the disjunctive particle to be exclusive. As I do not have relatable mother tongue data, it is not possible to conclude that these are cases of transfer, but it is still possible to make the conclusion that something seems to be influencing the comprehension of a third of the Japanese speakers of English, meaning that they form another comprehension of ‘or’ than that of the native speakers, i.e. a comprehension of ‘or’ as a signal that both people were killed but the reporter does not have any specific picture stored about the details of it in his/her mental stores.

104 The Chinese speakers of English:

The answers from the Chinese speakers of English are quite interesting for this text. As figure 8 below shows, the majority of them find that two people are killed, but the majority are far less convincing than especially the native speakers and the Russian speakers of English, and to some extent also more so than the Japanese speakers of English.

Figure 8 Overview of the answers from the Chinese speakers of English. Numbers of participants selecting a given answer are presented as percentages for ease of comparison across groups, as the different groups have slightly different numbers of participants.

As seen from the chart above, 39% of the Chinese speakers of English understand that only one person was killed. This is rather a lot. Especially since the Chinese speakers in general evaluate themselves to be very proficient in English (their self-evaluated average is 5,3 for reading skills).

Looking at the nine participants who selected only one victim, they concur with this pattern and all evaluate themselves between 4 and 7, with the average being 5,3, i.e. exactly the same as for the rest of the group. The rest of their demographic information does not reveal anything particular about them either. As with the Japanese speakers of English, I have not made a similar test in their mother tongue, which means I shall not make any conclusion that the 39% is a result of transfer from their Chinese mother tongue. However, it does show that there is no uniform understanding of the English disjunction with the Chinese participants, which could indicate that something is influencing their comprehension, and this is interesting.

39%

61%

61%

39%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

How many people were killed? How many killers were there?

Proportion of participants

Chinese speakers of English - overview of both questions

1 person 2 persons

105

There has been some work on cross-linguistic influence from L1 Chinese into L2 English, but as with the studies mentioned for the Japanese speakers of English, most seem to investigate disjunction in relation to negation discussing what takes scope over what. Furthermore, a lot of the studies relate to child acquisition of disjunction in different languages, and only indirectly relate to adult speakers. Crain (2012, p. 178ff) finds that adult native speakers of Chinese (Mandarin) are similar to native Japanese speakers in the sense that they interpret the Chinese disjunction ‘huozhe’ as meaning one or the other but not both, i.e. the exclusive interpretation, in simple sentences with negation. And, also similar to the Japanese interpretation of the Japanese disjunction described by Goro & Akiba (2004), Notley, Zhou, Jensen, & Crain (2012) found that the Chinese disjunctive particle ‘huozhe’ is interpreted as inclusive (i.e. similar to the English inclusive interpretation of ‘or’) when negation appears in a higher clause than the clause with the disjunction. But, as also discussed under the Japanese participants, it is unclear how this might relate to my data, if scope (or not) of disjunction may offer any explanation of the answers from those Chinese speakers of English who comprehended that only one person was killed.

In relation to the second question, the Chinese speakers of English resemble the rest of the groups, including the native speakers with a majority of the participants agreeing on one killer, and a minor group agreeing on two killers.

Summing up, it is interesting that so many of the Chinese participants display a preference for the exclusive interpretation of ‘or’ even though they assess themselves to be quite proficient within English. This indicates that their comprehension of meaning behind ‘or’ in this context is not the same as the native speakers, i.e. that they do not understand ‘or’ as a signal that both were killed, but reporter has no specific picture stored with details about this. Instead they comprehend ‘or’ as exclusive, i.e. in the ‘either…or’ sense.

The Russian speakers of English

As seen from figure 9 below, the Russian speakers of English resemble the native speakers in English quite a lot in that 92 % found there to be two victims. This is interesting as it contrasts with what Lorentzen (2016) found in her analysis of Russian and Danish use of conjunction and disjunction and what inspired this investigation to begin with.

106

Figure 9 Overview of the answers from the Russian speakers of English. Numbers of participants selecting a given answer are presented as percentages for ease of comparison across groups, as the different groups have slightly different numbers of participants.

Looking at the two Russian speakers of English who did not choose two victims, but only one, nothing seems to set them apart from the rest of the Russian participants and in terms of proficiency both evaluate themselves to be rather proficient in English when it comes to reading.

As this project does not include a similar test in their mother tongue, is not possible to say if the similarities in comprehension may be a case of positive transfer, i.e. if there is something in their native Russian which helps them make form this comprehension of the English ‘or’. As Lorentzen (2016, pp. 81-83) states, contrary to Danish (and to English) Russian has a large variety of different disjunctive particles or constructions available each with its own specific meaning and function. It may be possible that this system of disjunctive particles in some way helps them form a comprehension similar to the native speakers’. It is also possible that this similarity in comprehension is simply an indication that the Russian speakers of English are at a level of proficiency in terms of understanding English that reflect the native speakers of English. They understand that the English disjunction in the situation does not function as the Russian equivalent would (i.e. that only one of the two persons were killed), but instead they mirror the native speakers’ understanding, i.e. that ‘or’ in this text carries the information that two persons were indeed killed but that the speaker has no specific picture stored about this.

8%

76%

92%

24%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

How many people were killed? How many killers were there?

Proportion of participants

Russian speakers of English - overview of both questions

1 person 2 persons

107

Regarding the second question, the Russian speakers of English resemble all the other groups, including in particular the native speakers of English in that the vast majority (76%) choose one killer.

Summing up, the answers from the Russian speakers of English are interesting because they are quite similar in their comprehension to that of the native speakers. This is unlike what Lorentzen (2016) found in her thorough investigation of the use of disjunction and conjunction in translations between Russian and Danish, and perhaps also contrary to what I had expected, yet this does not make it any less interesting. It is important to notice that Lorentzen (2016) investigated the use of disjunction and not the comprehension of it. In other words, her investigation of the differences in this between Russian and Danish showed how differences in the communication process of reality-oriented and hearer-oriented languages affected something as basic and profound for language as the use conjunction and disjunction, which was otherwise considered to be universal across languages. However, it seems that at least for disjunction in this particular context, I cannot draw the same conclusion for the comprehension of it. In other words, though there are clearly differences in their use across Russian as reality-oriented language and Danish as hearer-oriented language, this small investigation found no real difference in the comprehension of the English disjunctive particle ‘or’ between native speakers of English and Russian speakers of English in this context.

Summing up on the comprehension of ‘or’

To briefly recapture the findings from the analysis of the comprehension of ‘or’, the answers of the participants showed several interesting items. First of all, as figure 10 below shows, all groups

108

show a majority for an inclusive reading of ‘or’, i.e. that two people were in fact killed.

Figure 10 Overview of the answers to the question of, how many people were killed, across all groups. Numbers of participants selecting a given answer are presented as percentages for ease of comparison across groups, as the different groups have slightly different numbers of participants.

However, it also showed that the native speakers were the only group to agree completely on this.

This means that for the native speakers of English the disjunction ‘or’ is fully grammaticalised, i.e. encompassed by the common code of the English language. We are able to explain this according to the communication process of the English language, as laid out in the theoretical foundation, by seeing the disjunctive particle ‘or’ as information to the hearer – this is indeed the main function of the grammar of a hearer-oriented language – and thus as a signal that the two combined elements (Mrs Hewitt and Dr Gott) are not stored in the same mental storage of the speaker, or the hearer for that matter, which implies that they are not killed at the same place and time. And this could of course also suggest that there could be two killers.

Secondly, the data showed that the majority of all the non-native speakers actually did resemble the native speakers in their understanding of the disjunctive particle ‘or’, but it did also highlight a few interesting differences. Most similar to the native speakers were the Russian speakers of English. This was interesting since it contrasted with the findings by Lorentzen (2016) on the use of disjunction and conjunction, which means that even though there are clear differences in the use of disjunction, I did not find clear differences in the comprehension of ‘or’ in this specific context. On the other hand, although the majority of the Japanese and Chinese speakers of English also understood the disjunction as inclusive, quite a few of especially the Chinese speakers of English interpreted it as exclusive, i.e. they selected that only one person was killed. This is

0%

29%

39%

8%

100%

71%

61%

92%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

England Japan China Russia

Proportion of participants

The different groups

How many people were killed?

1 person killed 2 persons killed