4.1 Subjectivity

Lyons (Reference Lyons1981, Reference Lyons, Jarvella and Klein1982, Reference Lyons1995) draws attention to the inadequacies of objectivist theories of semantics, where the speaker is little more than a neutral reporter of propositions derived from observation of the external world. These approaches ignore the role of subjectivity in language, that is, ‘the way in which natural languages, in their structure and normal manner of operation, provide for the locutionary agent's expression of himself and his own attitudes and beliefs’ (Lyons Reference Lyons, Jarvella and Klein1982: 102; see also Finegan Reference Finegan1995; Scheibman Reference Scheibman2002; and Traugott Reference Traugott1995 for comparable approaches). The speaker's self is not only the location of his beliefs, emotions, and attitudes, it is also ‘the product of the social and interpersonal roles that he has played in the past’, as well as ‘the role that he is playing in the context of utterance’ (Lyons Reference Lyons1981: 240).

Subjectivity, in the sense of the speaker's self expression, has been the focus of considerable interest in recent years (Athanasiadou, Canakis & Cornillie Reference Athanasiadou, Canakis and Cornillie2006; Stein & Wright Reference Stein and Wright1995; Traugott & Dasher Reference Traugott and Dasher2002). As these references indicate, subjectivity has many ramifications for the study of linguistic and semantic structure. Here, we wish to draw attention to three of its manifestations which will prove to be important for our study of causal connectives.

A first, and obvious, manifestation of subjectivity concerns the speaker's expression of her attitude towards the things and events she is talking about, as revealed by sentence adverbials such as regrettably, fortunately, and the like. Subjectivity also concerns the speaker's epistemic stance, that is, her degree of commitment to the truth of the expressed proposition (Traugott, Reference Traugott1989: 35; see also Traugott, Reference Traugott1995). On the basis of evidence available to her, in association with background assumptions concerning, for example, the likely course of events, a speaker may infer that a proposition may be true, can't be true, or must be true. Of special status in this connection are expressions in which the speaker explicitly talks about herself – her physical and emotional state, her past experiences, her decisions, intentions, wishes, and hopes, as well as her relations (present and past) with her interlocutors and with others. In contrast with statements concerning third parties and ‘objective’ states of affairs in general, a speaker's talk about herself comes with an epistemic guarantee, as it were, in the sense that the speaker is (or can consider herself to be) the ultimate authority on how she feels, what she wants, and what she has experienced. But even a plain statement of fact is imbued with a degree of subjectivity, to the extent that the speaker opts to present her particular construal of the situation (Langacker Reference Langacker1987: 128ff.). Moreover, every statement of fact is made against the speaker's presumption that the hearer is ignorant of this fact and would be interested to be informed about it. In this connection, we might more appropriately speak of intersubjectivity (Traugott in press; Traugott & Dasher Reference Traugott and Dasher2002: 20; Verhagen Reference Verhagen2005).

4.2 Cause and reason

Some semanticists refer to ‘cause’ as a semantic primitive, a predicate which enters into the semantic structure of certain lexical items and larger constructions (Goldberg Reference Goldberg1995; McCawley, Reference McCawley and McCawley1976). Others have emphasized the multifaceted and heterogeneous nature of causation, denying that it is a simple and undifferentiated, unanalysable concept (Lakoff & Johnson Reference Lakoff and Johnson1980: ch. 14; Talmy Reference Talmy1988). In the following, we seek to emphasize the inherently subjective and context-dependent nature of causal relations.

Consider a seemingly straightforward example of what looks like an objectively construed relation.

1. (9) The temperature has risen above zero. {Therefore, so, because of this, for this reason, etc.} the snow is melting.

Here we are presented with two situations, S1 and S2, whereby the occurrence of S1 caused S2 to come about. The relation may be explicated in terms of a counterfactual conditional: if S1 had not happened, S2 would not have happened. We may be confident of the validity of this kind of reasoning since it rests on natural laws and the necessity of certain outcomes, given a prior situation (Lewis Reference Lewis1973).

Example (9) exhibits what Halliday & Hasan (Reference Halliday and Hasan1976: 242) refer to as an external relation, that is, the relation holds in the outside world, independent of a human observer. Nevertheless, a certain degree of subjectivity inheres in our recognition even of the causal relation in (9). Although the relation is external and based on laws of nature, there is still involvement by the speaking subject. As Humean skeptics remind us, the truth of a natural law, and the necessity of cause–result relations, can never be decisively determined; our understanding of the workings of nature is contingent on previous observations and experience. Secondly, the validity of the relation is dependent on background assumptions, which, more often than not, are never explicitly stated (Meyer Reference Meyer2000). For example, both (9) and its counterfactual rewording presuppose normal atmospheric pressure. These presupposed conditions also play their part as causal factors in the occurrence of S2. A speaker selectively focuses on just one of a large (and possibly indeterminate) number of such factors. Thirdly, a speaker makes a choice regarding which links in the causal chain she wishes to highlight. There are events at the molecular level which intervene between the rise in the temperature and the melting of the snow, which are telescoped in (9). Alternatively, the speaker may zoom out, as it were, and declare that the melting of the snow was caused by the onset of spring, or by the warm westerly wind.

When we turn our attention to a wider range of examples we observe an increase in the subjective component. (In such cases we might be more inclined to describe the relation as one of reason, rather than cause.) There are two main dimensions of increasing subjectivity. Consider, first, the following examples.

1. (10)

   1. (a) The snow is melting because the temperature is rising.

   2. (b) The temperature is rising, because the snow is melting,

   3. (c) Do you want a lift, because I'm leaving now.

(10b) appears to reverse the causal relation in (10a), and indeed it makes no sense to assert that the melting of the snow causes the temperature to rise. (For a discussion of ‘backward’ causals, see Pit Reference Pit2006.) To be sure, (10b) presupposes the external relation. The speaker's knowledge that such a relation in general holds entitles her to draw an inference about the causing event, S1, given only her observation of the consequence, S2. The relation in (10b) is therefore epistemic (‘internal’, in Halliday & Hasan's terms), having to do with the reason, not why S1 is true, but why the speaker believes it is legitimate to believe S1, and hence to assert S1. Turning to (10c), the relation is no longer between two propositions. Rather, the second clause is presented as a factor which legitimizes the speaker's making the speech act of an offer.

The contrasts in (10) illustrate Sweetser's (1990: 76–9) distinction between the content, epistemic, and speech-act domains of causal relations. There is, however, another dimension to the subjectivity of causal relations. Compare the examples in (11).

1. (11)

   1. (a) The snow is melting because the temperature is rising.

   2. (b) She left the party because it was getting late.

Both examples have to do with events in the external world. But whereas the relation in (11a) concerns the workings of a natural law, in (11b) the relation is mediated by the decision of a conscious agent. On what is perhaps the most natural interpretation of (11b), the woman became aware that it was getting late (S1) and decided, on that basis, that it was necessary, or appropriate, that she should leave (S2). The subjectivity of the relation is therefore not primarily located with the speaker but with the decision-making agent. Even so, we can identify increasing degrees of speaker-centred subjectivity. For example, the speaker may simply be reporting the causal relation entertained by the protagonist: ‘She told me why she left, namely, because it was getting late.’ On another, and more intrinsically subjective interpretation, the speaker empathizes with the protagonist, reconstructing her mental processes. There is a third possibility, which imputes no decision-making to the protagonist.Footnote ⁹ Suppose the woman in question habitually leaves parties early. For the speaker, the relation between S1 and S2 would be part of the normal course of events. In this case it would be the speaker, not the protagonist, who establishes the relation between the events.

These matters have been discussed by Pander Maat & Sanders (Reference Pander Maat and Sanders2000, Reference Pander Maat and Sanders2001) in terms of the conceptual distance between the speaker and a ‘Subject-of-Consciousness’ (SoC).Footnote ¹⁰ Concerning (11b), on the first of the interpretations discussed above, the SoC is the woman, who decided to leave when she observed that it was getting late. The conceptual distance between speaker and SoC is maximal. Conceptual distance decreases (and subjectivity increases) in case a speaker reports her own volitional actions (12a). Conceptual distance is minimized if the speaker refers to her own current circumstances and current decision-making (12b).

1. (12)

   1. (a) I left the party because it was getting late.

   2. (b) I'm leaving now because it's getting late.

4.3 Information structure

Dancygier & Sweetser (Reference Dancygier and Sweetser2005) draw a broad distinction between two kinds of causal conjunctions. We can refer to these as because-type conjunctions and since-type conjunctions. A ‘test’ for the two types has to do with whether the causal relation is compatible with various focusing strategies. Because-type conjunctions permit focusing, negation, and contrast; since-type conjunctions do not. Example (11b) permits the variations in (13). Analogous sentences with since are not possible (14).

1. (13)

   1. (a) It was because it was getting late that she left the party.

   2. (b) She didn't leave the party because it was getting late but because she was feeling sick.

2. (14)

   1. (a) *It was since it was getting late that she left the party.

   2. (b) *She didn't leave the party since it was getting late but since she was feeling sick.

Conjunctions which pattern like since include as, for, and, of course, seeing (that) and seeing as though. Because-type connectives include due to (the fact that), for the reason that, and in view of (the fact that).

As a matter of fact, the term ‘because-type conjunction’ is not ideal, in that because can sometimes pattern with the since-type conjunctions. Consider the following contrast (cf. Dancygier & Sweetser Reference Dancygier and Sweetser2005: 180–3). In (15a) – written without the comma and spoken as a single intonation unit – the speaker gives the reason why the person left. In (15b) – written with a comma and spoken as two intonation units – the speaker asserts that the person left, and then states the reason why this event happened, or why she feels entitled to state that it happened.

1. (15)

   1. (a) She left the party because it was getting late.

   2. (b) She left the party, because it was getting late.

The contrast becomes clearer if the sentences are embedded under an epistemic verb:

1. (16)

   1. (a) I believe she left the party because it was getting late.

   2. (b) I believe she left the party, because it was getting late.

In (16a) the reason is part of the speaker's belief; in (16b) the fact that it was getting late is adduced in support of the speaker's belief. The commaless version (16a) exemplifies a because-type relation, (16b) exemplifies a since-type relation.Footnote ¹¹

De Smet & Verstraete (Reference de Smet and Verstraete2006: 383) propose for because-type causals the semantic structure (I assert [X is caused by Y]). The formula captures the status of the causal clause as part of a complex proposition. For since-type constructions they suggest the structure (I assert X, I say it is caused by Y). The formula highlights the illocutionary independence of the causal clause. The incorporation of the I say component also recognizes the inherent subjectivity of a since-type conjunction.

De Smet & Verstraete's formulas have the causal clause in second position. Their explications should not, of course, be taken as mapping directly onto the linear sequence of clauses. Indeed, for many of the connectives, both initial and final position of the causal clause are fully acceptable (a fact which may be verified on the examples cited in this article). Nevertheless, differences in the preferred order for the two types might be expected, due to the different information structures associated with them.

In a because-type construction, as in (15a), the main clause is likely to contain presupposed, or given information, the speaker's point in making the utterance being to state the reason why the state of affairs came about (Dancygier & Sweetser Reference Dancygier and Sweetser2000: 135). On the presumption that given information is likely to be presented first (Prince Reference Prince and Cole1980), we would expect because-type clauses to occur in sentence-final position. In a since-type construction, on the other hand, the reason clause is typically taken as presupposed. The speaker first lays out presupposed information and then builds on this with the assertion, query, or whatever, that she is intent on making (Verhagen Reference Verhagen2005: 170). The reason clause might thus be expected to occur in initial position. These ordering preferences have been confirmed in several studies (Altenberg Reference Altenberg1984; Diessel Reference Diessel2001, Reference Diessel2005; Ford, Reference Ford1993). Deviations from the preferred orders can, of course, be expected. A since-clause might be postposed, almost as an afterthought, in case the speaker seeks to justify, post hoc, the content of the main clause.

We have said that the information introduced by a since-type conjunction is likely to be presupposed, as when the speaker takes up material already introduced into the discourse, summarizing it or alluding to it. Alternatively, the speaker refers to generally accepted facts, which might reasonably be taken to be common knowledge. We should also bear in mind the possibility that the speaker insinuates, for rhetorical effect, that the content of the reason clause is presupposed and therefore beyond dispute. Especially this last possibility can make it difficult in practice to assess whether the information in a clause really is presupposed or is newly asserted. As a result, the claim that since-type conjunctions introduce presupposed information might be difficult to confirm empirically.

There are, however, cases where this interpretative uncertainty does not arise (cf. Dancygier & Sweetser Reference Dancygier and Sweetser2000: 129–30). One case is where the information in the since-clause pertains to the hearer, or to an experience previously shared by hearer and speaker. Such information is, presumably, accessible to the hearer; the speaker is, at most, simply reminding him of it. The following (BNC) examples illustrate:

1. (17)

   1. (a) Since you're moving out anyway, I don't see why you worry.

   2. (b) ‘Since you've already saved my life once, I'm sure I'm quite safe!’, she retorted at last.

   3. (c) Since you ask me sir I can't think of a single reason.

A second case is when the speaker refers to some facet of herself, such as her thoughts, feelings, or physical state, as the basis for some further remark, as in the following (BNC) examples:

1. (18)

   1. (a) Since I am a short, plump, Asian young woman, with glasses and shortish hair, I've never found an advert with a woman looking anything like me.

   2. (b) Since I've got no idea why you wanted to see me, you'd better tell me what's on your mind.

   3. (c) Since I know who you are and you obviously know me, there's no need for introductions.

Note that the information in the since-clauses may actually be ‘new’ to the hearer. From the speaker's perspective, however, the information is beyond dispute; the speaker is, after all, the ultimate authority on matters pertaining to herself.

4.4 The differentiation of causal relations

A striking fact about English (and no doubt other languages) is the large number of resources that are available for the expression of relations of cause/reason. In addition to since and because, causal connectives, in English, include such items as so, as, for, therefore, consequently, that's why, for this reason, due to the fact that, and many more. Some of these, such as because, can be regarded as ‘pure’ causals, while others, such as since and as, are polysemous between causal and diachronically older noncausal readings. In addition, there are quite a few resources which, while not inherently causal, may conventionally or conversationally implicate causality. Consider the temporal relations in the following examples (Traugott & König Reference Traugott and König1991: 194–7). Probably, (19a) would be interpreted to mean that hearing the lecture caused us to feel inspired.

1. (19)

   1. (a) After we heard the lecture we felt greatly inspired.

   2. (b) I can't sleep now that I'm alone.

   3. (c) The minute John joined our team, things started to go wrong.

Consider also the interpretation of participial clauses, as in the following BNC examples.

1. (20)

   1. (a) Having known her for less than two months, he was not yet confident of any strong hold on her affections.

   2. (b) The weather being as it is, she'll no doubt remain there overnight.

   3. (c) the easily accessible position of the Malverns near big population centres make them a popular target for picnickers and day trippers . . . This being the case, summer weekends are not the best time to visit.

   4. (d) Given that there are several ways of doing a calculation, we can ask which of them, if any, is analogous to the way the brain performs the same calculation.

The information in the participial clauses is backgrounded and presupposed, and in the above examples is likely to be construed as justifying or legitimizing the assertions in the main clauses. Interestingly, in (20d), the presupposed information is explicitly characterized as ‘given’.

The various connectives available to an English speaker differ in a number of ways. These include the sequence in which S1 and S2 must be or are typically stated, which of the two clauses is prefaced by a connective, whether the causal relation can be focused, and whether the connective has the status of a subordinating conjunction or is more adverbial in nature. Semantically, as we have seen, connectives are differentiated according to the external or internal status of the causal relation, also with respect to various aspects of subjectivity. The large number of resources available to an English speaker very likely reflects the semantic differentiation of the causal relation itself.

In spite, then, of our earlier remarks that seeing as though appeared to have roughly the same semantic value as other expressions, such as since and seeing that, we might expect these connectives to be associated with somewhat different, or at least only partially overlapping, conditions of use. To investigate this matter we proceeded to draw up a semantic profile of seeing as though. We performed parallel analyses on two other connectives for purposes of comparison and contrast, in the hope of pinpointing specific factors which might motivate the use of seeing as though. Since was selected as it is probably the most general member of the since-type category. Seeing that was selected because of its sharing of lexical material with seeing as though.

For this purpose we compiled a small database consisting of 50 instances each of seeing as though, since, and seeing that. As already mentioned, seeing as though is very sparsely represented in the BNC. We therefore assembled examples on the basis of Google search results. Rather than taking the first 50 hits, we sampled Google returns at intervals of 100, selecting the first 10 results at each interval. In most cases we retrieved the web page and included the surrounding context in the database. Repeated hits (which were very few) were ignored, as were fragmentary examples where the syntactic affiliation of the seeing as though clause was not clear. Where a web page contained more than one instance of the expression we included only the first. In order to safeguard comparability of data, examples of seeing that and since were also sourced from the Web using the procedure described above. Inappropriate hits (e.g. temporal since) were of course ignored.

5.1 Clausal subjects

A first, general indication of the subjectivity of a causal construction can be gained by considering what each of the two clauses is ‘about’. We take it that clauses which are about the speaker, her opinions, thoughts, and decisions, her recent experiences or her past history, are inherently more subjective than reports about third parties or descriptions of ‘objective’ states of affairs (Scheibman, Reference Scheibman2002).

A simple measure of what a clause is about is to consider the category of the grammatical subject, whether first person (I or we), second person, or, if third person, whether human or other (animate or abstract). For this purpose we ignored the contribution of ‘main clause’ expressions of the speaker's epistemic stance; thus, in (21a), we took the subject of I think this place will get some good business to be ‘this place’, on the grounds that the expression makes an assertion about the place, not about the speaker (cf. Diessel & Tomasello Reference Diessel and Tomasello2001; Halliday Reference Halliday1985: 333; Verhagen Reference Verhagen2005).Footnote ¹² We also ignored the effects of focusing and extraposition. Thus, in (21b), we took the subject of the focused clause to be he.

1. (21)

   1. (a) It's only a small store, but I think this place will get some good business, especially seeing as though the over-priced Debacle is directly opposite.

   2. (b) thats what i think he will do this week after around 20 minutes of the game (seeing as though he has mellars coming off the bench).

Results are summarized in table 1. Note that in almost half the seeing as though data the consequent clause is about the speaker. The figures for the reason clause are somewhat lower, yet markedly higher than for the two other connectives. On this measure, seeing as though constructions turn out to be more speaker-centred than the other types. The data suggest that seeing that is the least speaker-centred of the three expressions, with since having an intermediate value. On the other hand, since would appear to be the preferred option when a causal construction is about nonhuman (including inanimate and abstract) entities.

Table 1. Subject of reason clause (r/c) and consequent clause (c/c)

Note: Differences in the ‘speaker-centredness’ of three connectives are extremely significant; χ² (2) = 14.600, p = 0.0007.

5.2 The content of the reason clause

We noted earlier (section 4.2) that subordinated reason clauses are likely to contain information which the speaker assumes as given and which is invoked to justify, or explain, the contents of the consequent clause.

Initially, we had intended to classify the contents of the reason clauses according to their information status, whether common ground or newly asserted. We quickly discovered that decisions in this regard were often variable and unstable. We therefore turned to more ‘objective’ criteria of analysis.

First, we identified those reason clauses whose content pertains to some aspect of the speaker herself, such as her mental or physical state, her recent experiences, or her past history. Although the content of such clauses might well constitute ‘new’ information to the hearer (the hearer might be ignorant of the speaker's age or of her activities in the recent past), the information is, for the speaker, irrefutable; as such, it is able to constitute an unassailable basis for the content of the main clause.

About one-third of the seeing as though clauses (table 2) fell into the speaker-centred category. The following examples illustrate:

Table 2. The source of the information in the reason clause: (a) the speaker's mental or physical state, (b) some facet of current situation, (c) matters not covered by these categories

Note: Differences in the speaker-relatedness of the three connectives are significant: χ² (2) = 9.172, p = 0.0102.

1. (22)

   1. (a) now all this may seem childish from another viewpoint and i assure you it is, but seeing as though i cannot post on the forum they have free range to slander me without reprisal.

   2. (b) The financial person said if we wanted to carry on paying the £40 a month, we could have the money in years to come, which by the way should have been on the 4th of July this year. Seeing as though we hadn't missed the £40 we decided to carry on with it for 8 more years, at the end we should be getting £6,000.

   3. (c) I don't remember that series seeing as though I was only 3 years old at the time.

The seeing that data contain only a couple of comparable examples, of which (23a) is one. There were, in addition, two seeing that examples which report another's self-reference, as in (23b).

1. (23)

   1. (a) I'm not saying I'm different or better than these people – I do the corporate stuff every day too. But, seeing that I do run K10k with Toke, I can honestly say that we're at least trying to be different – and trying to feature things from people who don't mind breaking all the rules.

   2. (b) Wherefore will ye die, O ye house of Israel, seeing that I have no pleasure in the death of him that dieth, saith the Lord God?

Since-clauses, on this measure, again occupy an intermediate position, with five examples where the speaker appeals to her present situation and three which refer to her past history.

1. (24)

   1. (a) Also, since I know that the author of the strip hangs around S_D – have we corrupted another soul?!

   2. (b) . . . since I have encountered apparent skepticism from certain individuals who were only aware of parts of the story, I have decided that it is best to reveal pretty much everything.

A second category concerned those reason clauses which pertain to some aspect of the discourse context, such as the behaviour, attitude, or recent remarks of the listener, matters concerning persons or things currently under discussion, or aspects of the extralinguistic context (such as the weather or the approaching holidays). The information in these clauses might reasonably be regarded as given, or at least easily accessible or retrievable, from the perspective of the speaker and other speech act participants. However, we prefer to emphasize the status of this information as being anchored in the speech act situation.

On this measure the three connectives are roughly comparable (though seeing that scores a little lower than the other two). In (25a), the reason clause refers to some feature of a display available to the addressee and which forms the basis for the speaker's instruction; (25b) refers to some aspect of the current discourse; while (25c) draws attention to some feature of the environment.

1. (25)

   1. (a) Park yourself in the Piper Malibu at Brisbane International Airport (YBBN), and find your GPS. The route we will be taking today is CG ENTRA YSSY. Seeing as though the first waypoint on the list is CG (it's a VOR to save you some time), press the VOR button on the GPS and enter ‘CG’. It should say ‘Coolangatta’ on the GPS screen now.

   2. (b) Seeing as though this is a broad topic, we went with questions instead of key messages.

   3. (c) It's only a small store, but I think this place will get some good business, especially seeing as though the over-priced Debacle is directly opposite.

Here are some comparable examples with seeing that and since.

1. (26)

   1. (a) Seeing that the winter holidays are coming up, what would all of you like to get?

   2. (b) If either of the first two possible courses of action had been taken, it's a safe bet that the probability of this turning into a viable project are almost nil. By remaining flexible and seeing that there are many different ways this situation could play out, you save yourself a great deal of self-induced stress.

2. (27)

   1. (a) Given you have a browser capable of displaying flash and a working internet connection (assumed, since you're reading this), you can check out a Mii maker.

   2. (b) since this survey is now restricting itself, in the modern period, almost completely to Czech-language writing, we must really only mention him here in passing.

For the sake of comparison, we cite some of the examples which were included in the category ‘other’ of table 2, where the reason clause pertained neither to the speaker nor to features of the current discourse.

1. (28)

   1. (a) Seeing as though veterinary and health services are generally weak, the avian flu has already resulted in 6 fatal cases since 2005 in Cambodia.

   2. (b) Because we've decided that huge sums of money are seemingly miraculous, and seeing as though the chances of winning the lottery is about 1 in 25 million, we could suppose that our chances of being part of a miracle is about 0.000004%.

2. (29)

   1. (a) Seeing that lawyers have squeezed every drop of live [sic] out of doctors, they need a new target.

   2. (b) Uesbeck and Topp studied the feeding performance in Oniscus asellus; it is an important subject which is now studied by several authors, seeing that isopods are a common component in the soil fauna and are known to feed on a variety of decaying plant material . . .

3. (30)

   1. (a) Starting the series in 1987 was an unfortunate choice, since the late 1980s and early 1990s was a difficult time for the world's poor.

   2. (b) You will be able to find cheaper inspectors, but some may have very little experience since state licensing does not require any actual home inspection experience.

In many of these cases, to be sure, the content of the clause might plausibly be regarded as constituting common knowledge shared by speaker and hearer, and presented by the speaker to bolster the content of the consequent clause. However, as noted above, a decision regarding the actual status of the information (as shared or new) is a tricky matter and difficult to decide conclusively. The point about these examples, rather, is that the content is not specifically tied to the particular circumstances of the speech act. To this extent, the examples are anchored in the speech situation to a lesser degree than those cited in (22)–(27).

5.3 The status of the consequent clause

The dominant function of consequent clauses in seeing as though constructions is to express the speaker's epistemic stance and/or the speaker's attitude. Almost two-thirds (30 examples) fell into this category. The clauses expressed the speaker's opinion (8 examples), an evaluation (6), a decision (6), or a suggestion (2). A further 8 examples contained an inference drawn by the speaker, often in association with a modal element (seem, must, etc.). As will be apparent from the following examples, the speaker's opinion, inference, etc. arose from considerations raised in the reason clause.

1. (31)

   1. (a) Merima and I are livid with the fact that we have to leave in a week. Our petition for citizenship will not go in the direction we hoped. I guess I will finally have to really learn the language, seeing as though the only way I will be able to become a citizen, is to get married.

   2. (b) According to Microsoft one of the major benefits of using WPF is the improved designer-developer workflow. But seeing as though WPF is still in its infancy, it's somewhat confusing as to how exactly designers should be working and in what tools.

   3. (c) I decided to have a look around for some clothes, seeing as though they had a special on ladies wear.

   4. (d) Seeing as though every ZVM is still under warranty, Im sure you could get yourself a new one.

In addition, there was one example where the speaker issues an instruction to the hearer (25a), and three examples where the speaker raises a query, as in the following:

1. (32)

   1. (a) Seeing as though Exetel resell Optus, does that mean i can get Optus DSL Direct in October?

   2. (b) I've heard that Switzerland and the US both have possible surprise teams, so good luck to them! And of course, even Roony-less, England is always a big contender, so I hope they do well. Well – seeing as though I support 5 teams, is it possible that 2 out of five make it to the end?? We'll see!!

In the remaining 16 examples in the seeing as though database, the main clause did not contain any overt expression of the speaker's epistemic stance, nor any modal or evaluative elements. Rather, the clause merely stated a fact, reported an event, or described a situation, with the reason clause offering a justification, explanation, or background information relevant to the assertion.

1. (33)

   1. (a) Dad and i headed out to the lakehouse at like 9:45 seeing as though i dont eat breakfast. i was already packed and ready to go with the iPod so we were off.

   2. (b) Seeing as though my parents now live in the United States (I'm gonna rope em into buying me a PSP while they are there and post it back ☺) and I have been left in this huge ass home by myself, a mate of mine has moved in and brought his PS2 with him.

2. (34)

   1. (a) The River Neckinger flows into the Thames at St Saviour's. The name means ‘devil's necktie’ – or hangman's noose, seeing as though the dock is opposite an ancient execution site for Thames pirates.

   2. (b) The avian flu in Cambodia not only threatens 75% of the population relying on subsistence farming, but also poses an immediate danger to those susceptible to this possible pandemic. Seeing as though veterinary and health services are generally weak, the avian flu has already resulted in 6 fatal cases since 2005 in Cambodia.

The examples in (34) are especially interesting, in that they are the only examples in the seeing as though database which might possibly be said to involve an ‘objective’ causal relation, between circumstances that hold in the external world and which do not involve human agents. Thus, (34a) explains the name by reference to the location, while (34b) relates the mortality rate to the state of the health services. Nevertheless, we do perceive a subjective nuance in these examples. In (34a) the speaker seems to be suggesting that the name is a suitable one, given the historical background. The speaker, in other words, is offering an implicit evaluation of the appropriateness of the name and the validity of its supposed etymology. In (34b) we also detect the speaker's attitude, especially in the use of the adverbial already. It is as if the speaker is not really surprised that as many as six people have died, given the facts already stated.

Turning to the seeing that data, we again find a substantial number of examples where the main clause has a strongly subjective element. These comprised evaluations (5), opinions (3), decisions (1), suggestions (1), and queries (4).

1. (35)

   1. (a) Seeing that human destructive activity is accelerating I am in favor of Nature bringing a swift end to the oil industry.

   2. (b) Seeing that a lot of these reviews are before the digitaldom that we live in now, I wanted to chime in about this lens and the Canon 10D.

   3. (c) Seeing that the ‘blogosphere’ is so important, apparently, according to everything I've read, I've been remarkably remiss.

There were, however, several examples of inferences, evaluations, etc. which lacked the strongly speaker-centred subjectivity of the above examples. In the examples in (36) the speaker might well endorse the inferences and evaluations. These, however, are presented ‘impersonally’, and not due specifically to the speaker, but rather are attributed to some generic or bureaucratic entity.

1. (36)

   1. (a) The day is coming when all the religions of the world will unite, for in principle they are one already. There is no need for division, seeing that it is only the outward forms that separate them.

   2. (b) Seeing that the needs for common services will be based on the needs of a given simulation community represented within a federation, it is felt that it is unlikely that a set of common services that would be encompassing enough to meet the anticipated needs of all users could be developed.

   3. (c) People dealing with probate matters in the state of Maine, should be aware of a few important factors which ultimately contribute to the outcome of a given probate experience. The first has to do with understanding the ways that probate procedures work in Maine. This is particularly important seeing that the only way to get through the Maine state probate process is by adhering to the standardized procedures.

Other examples present ‘logical’ inferences; these pertain to academic or philosophical–theological registers, registers which are not represented at all in the seeing as though data.

1. (37)

   1. (a) Seeing that it is the nature of uniform circular motion to sweep out equal angles at the center in equal times, and given that these angles at the center belong to similar circumferences, which are, nevertheless, proportional to their radii, it thus follows that when the planet passes through point E belonging to the circle whose semi-diameter is SE, then its motion will be slowest.

   2. (b) Death, therefore, the most awful of evils, is nothing to us, seeing that, when we are, death is not come, and, when death is come, we are not.

Another important category, and one which in our data is associated exclusively with seeing that, is found in the context of third-person narratives. A protagonist becomes aware of a state of affairs described in the seeing that clause, and, as a consequence, acts in a certain way. There were in all 14 such examples in our database.

1. (38)

   1. (a) Vario, seeing that his command was not obeyed, became furious, and demanded the instant return of the boat.

   2. (b) Six of the negroes, seeing that their doom was sealed, killed themselves.

   3. (c) Seeing that they are approaching a crater about the size of a football field and covered with large rocks, Armstrong takes over manual control and steers the craft to a smoother spot.

With respect to the since data, we again find a substantial number of examples where the consequent clause is firmly anchored in the speaker's here and now.

1. (39)

   1. (a) Since I feel that lies have been told, and since I have encountered apparent skepticism from certain individuals who were only aware of parts of the story, I have decided that it is best to reveal pretty much everything

   2. (b) Since folks have so much trouble with ‘racist’, why don't we try ‘bigot’?

   3. (c) Since it won't stop raining . . . I spent the afternoon finishing up Season 3 of Oz.

However, on the whole, the since data are characterized by a somewhat lower degree of speaker involvement than the seeing as though data and most of the seeing that data. For example, since is used to report other people's inferences:

1. (40)

   1. (a) Analysts say this is significant since it's the first time the stock's been above $20 since May 2002.

   2. (b) Jenkins says since al-Qaida is now an ideology, the removal of Osama bin Laden would have less effect on the whole terrorist enterprise now than it would have had four or five years ago.

Here there is a divergence between the speaker and the Subject-of-Consciousness responsible for the inference (though, to be sure, the speaker may well sympathize with the view she reports). Notable, once again, are several examples which attribute the inferencing process to a ‘generic’ or nonspecific agent rather than to any identifiable individual. Thus, the reasoning in the following is taken to be a matter of ‘common sense’, to be generally accepted.

1. (41)

   1. (a) Since 40 percent of the world's food comes from irrigated land, water shortages can quickly translate into food shortages.

   2. (b) The attitude that women are responsible for adultery and rape is a catalyst for both, since it not only erodes the thin self-restraint of potential male predators, but also reduces the likelihood that women will report the offenders.

In yet other examples, we can attribute the inference (as well as the evaluative judgement) to the speaker; however, the subject matter is of a nonpersonal nature, having to do with situations in which the speaker is not a participant.

1. (42)

   1. (a) Starting the series in 1987 was an unfortunate choice, since the late 1980s and early 1990s was a difficult time for the world's poor.

   2. (b) Since these Syrian historical flags pages contain basically a compilation of several contributions, there is no way to tell out who reported the 1932–1946 Syrian flag with stars pointing downward.

Not surprisingly, these examples are found in more formal registers (book reviews, political commentaries), rather than the highly personalized blogs, chatrooms, and the like which are typical of most of the seeing as though data.

5.4 Locus of causal relation

Here we considered the ‘locus’ of the causal relation, that is, the agent who is responsible for establishing the relation between the two clauses, whether the speaker, a protagonist other than the speaking subject, or a generic, unnamed third party.

As our earlier discussion in section 4.2 showed, alternative interpretations of causal constructions are often possible. She left the party because it was getting late (11b) most likely reports a causal relation established by the protagonist. Even here, though, a more subjective interpretation can be envisaged, in which it is the speaker who adduces the reason for the person's departure. In evaluating this aspect of our data we chose what we took to be the most plausible or natural reading of the examples; we did not attempt to coerce a more subjective interpretation.

The results (table 3) confirm the trends reported so far. Seeing as though is used almost exclusively to express a relation established by the speaker, only rarely to state an ‘objective’ relation attributable to a generic third party, as in (34), or to report a relation established by a third-person protagonist. Thus, (43) is most plausibly taken to attribute the reason relation to Biggie, rather than to the speaker (though the latter interpretation, to be sure, cannot be ruled out).

Table 3. The locus of the causal relation between the reason and the consequent clauses

Note: Differences in the speaker-locus of the three connectives misses significance: χ² (2) = 4.971, p = 0.0833. However, compared to seeing that and since combined, the speaker-locus of seeing as though is significant: χ² (1) = 4.971, p = 0.0258.

1. (43) Biggie wanted to start a clothing line that would cater to ‘Big’ men like himself, seeing as though it was hard to find the quality of clothing he admired.

The seeing that data contained a significant number of examples where the relation is attributed to a named protagonist, as in the narrative examples cited in (38). The since data, on the other hand, scored highest for a generic, impersonal, or bureaucratic locus.

5.5 Expression of speaker stance

Next we considered overt markers of the speaker's epistemic stance, such as I think, I suppose, it seems to me, and the like, whether in the consequent clause or in the reason clause (cf. Mortelmans Reference Mortelmans2006). Subcategories concerned the reported epistemic stance of a protagonist or the presumed stance of the addressee. We also counted the number of other stance expressions, such as presumably, probably, really, and the like. We included here epistemic uses of modals such as must, can't, and will, as well as expressions of speaker attitude. Needless to say, one and the same example could contain several instances of these categories.

Seeing as though scored highest on both of these measures, confirming the high degree of subjectivity associated with this connective; since registered the lowest values (table 4).

Table 4. Incidence of overt expressions of stance (e.g. I think) and other stance expressions (e.g. probably)

Note: Differences between the three connectives with respect to expressions of overt epistemic stance are significant: χ² (2) = 7.875, p = 0.195, as are the differences with respect to all stance expressions: χ² (2) = 8.107, p = 0.0174.

5.6 Sequence

For all three connectives there was a preference for the reason clause to occur before the consequent clause. The preference is strongest for seeing that (35 examples of initial reason clauses), weakest for since (27 examples), with seeing as though occupying an intermediate position (30 examples). The differences are not significant: χ² (2) = 1.065, p = 0.5879. The overall preference for initial position is consistent with the presumed information structure of since-type connectives (section 4.3). In about a third of the cases, however, speakers justified the content of the main clause in a postponed reason clause.

5.7 Conspectus

Our discussion has shown that seeing as though constructions are – in Scheibman's (2002: 167) phrase – strongly ‘anchored to the speaker’. The reason clause typically states some fact about the speaker, while the consequent clause typically expresses the speaker's opinion, evaluation, decision, etc. Also characteristic of the data is the high incidence of overt expressions of speaker stance and speaker attitude.

Elements of subjectivity are not, of course, absent from the two other data sets we analysed. Indeed, one conclusion could be that there is considerable overlap in the usage range of the three expressions. Some of the examples with since and seeing that would, it seems to us, be quite acceptable with seeing as though, particularly when they exhibit the various kinds of subjectivity associated with seeing as though. Conversely, since could probably serve as an adequate replacement for many of the examples with the other two connectives. Nevertheless, the data do point to a certain degree of specialization. Since seems to be preferred when the speaker takes a somewhat detached perspective on the relation, reporting other people's inferences or attributing general validity to the relation. Seeing that, on the other hand, has a specialized use in narrative reports, presenting the perspective of one of the protagonists.

In light of these observations, it is instructive to return to the student's email which triggered this study. Replacing the student's seeing as though with since would, we feel, dramatically change the tone of the communication. The student would no longer be making a tentative suggestion, but a peremptory request.

1. (44) Dr X could supervise my thesis, {seeing as though/since} she is a native speaker of German.

The difference, we suggest, arises from the status of the relation between Dr X's being a native speaker of German and her suitability to supervise a thesis on the German language. Seeing as though attributes the relation to the writer and her construal of the situation, while since suggests that the relation has more general validity. In the first case, the validity of the causal relation is contestable; in the second case it is not.Footnote ¹³