2.1. cognitive approaches to OCD

The Cognitive-Behavioral Model (CBM; Salkovskis, Reference Salkovskis1985) is the most recognized approach to cognition in OCD. CBM maintains that intrusive thoughts are widely experienced also by by people without OCD (OCCWG, 1997; Rachman & de Silva, Reference Rachman and de Silva1978; Salkovskis, Reference Salkovskis1989), and that what sets individuals with OCD apart is their appraisal of such intrusions. According to CBM, obsessions and compulsions result from dysfunctional beliefs. The OCCWG proposes that six belief domains are central to OCD and that these influence the individuals’ appraisal of intrusive thoughts: inflated responsibility, “the idea that one has power which is pivotal to bring about or prevent subjectively crucial negative outcomes” (1997, p. 677); overimportance of thoughts, “the belief that the mere presence of a thought indicates that it is important” (1997, p. 678); the importance of controlling one’s thoughts, the notion that it is “possible and desirable” to control thoughts (1997, p. 678); overestimation of threat, the exaggerated belief that harm is very likely; intolerance to uncertainty, a belief that certainty is necessary and change/ambiguity can be very difficult to manage (1997, p. 678); and perfectionism, the idea that “there is a perfect solution to every problem” and that this solution must be found (1997, p. 678). CBM has been substantiated through clinical and experimental research and the model has also yielded successful approaches to the treatment of OCD.

Proponents of a more recent cognitive model of OCD, the Inference-Based Approach (IBA), argue that some of CBM’s cognitive belief domains “are more reminiscent of process variables or biases rather than particular beliefs” (O’Connor et al., Reference O’Connor, Aardema and Pélissier2005, p. 7). IBA focuses on ‘patterns in reasoning’ (cognitive processes) as opposed to the ‘mental content’ upon which such processes operate (2005, p. 9). It is argued that “process variables operate independently from specific mental content, and may apply to a wide variety of mental contents” (2005, p. 7). In line with IBA, this study maintains that distinguishing between cognitive operations and the conceptual content upon which they operate facilitates more nuanced descriptions and explanations of conceptualization. A principled distinction is made in this study between cognitive processes (construal) and conceptual structures (LOC; Paradis, Reference Paradis2005). This allows for nuanced descriptions of meanings that involve systematic matching of different types of conceptual structure (configurations and content structures) and their mapping to language through construal (Section 4).

3.1. conditional constructions and ocd

Conditional language and hypothetical reasoning (e.g., if P then Q and what if P) have been described as features of OCD. For instance, Shafran notes that people with OCD use “a ‘what if … ?’ questioning style to potential problematic features of their life” (2005, p. 237, italics added). If conditionals are also used to describe features of OCD, as when Abramowitz and Jacoby write that “motivation to resist [an obsession] is activated by the fear that if action is not taken, disastrous consequences will likely occur” (2014, p. 222, italics added). In the OCCWG’s discussion of cognitive assessment of OCD an if construction is likewise used to suggest that “it may be possible to include sentence fragments with portions of the sentence to be completed by the respondent (e.g., ‘If I have the thought that _____ and don’t do _____ then _____ will happen’)” (1997, pp. 674–5). The data sample includes 156 instances of if. The testimonial with the most instances (OCD 3) contains 26 uses of if and the one with the fewest (OCD 7) includes 6 instances. What if appears 20 times in the data and in 5 of the testimonials. The testimonial with the most instances of what if contains 8 (OCD 3). Unless appears 8 times in the data and 7 of those instances are in one testimonial (OCD 11).

If constructions have broad use potential that overlaps with other constructions, and expression of conditionality is not limited to if constructions. Dancygier and Sweetser (Reference Dancygier and Sweetser2005) provide an in-depth treatment of conditionality in English in which they adopt a mental space approach (Fauconnier, Reference Fauconnier1994). They write that the construction of conditional meaning “involves setting up a mental space (in the case of if-conditionals, this is the job of the if-clause), and requesting construal of something (in if-conditionals, the then clause or main clause) within that space” (Dancygier & Sweetser, Reference Dancygier and Sweetser2005, p. 18). DST offers a compatible account wherein the if-clause is described as setting up a deictic space through transformation of S’s base reality space, according to which the then-clause / main clause can be construed (Section 5). While if itself is treated in the literature as a ‘space builder’ (e.g., Fauconnier, Reference Fauconnier1994, Reference Fauconnier1997) and as a ‘cognitive operator’ (Chilton, Reference Chilton2014), it is acknowledged that if does not effect conditional construal on its own. The meanings evoked by if constructions are contingent on, among other things, “the meaning of if, the tense of the verb, the semantics of the particular verb, expected intentions of the speaker and other contextual factors” (Chilton, Reference Chilton2014, p. 160). Tense in particular plays a significant role in the evocation of conditional meaning (Dancygier, Reference Dancygier1998; Dancygier & Sweetser, Reference Dancygier and Sweetser2005; Fauconnier, Reference Fauconnier1994).

7.1. counterfactual conditionals

Counterfactual conditionals evoke contrast between a scenario construed by S as real (modally proximal) and another one construed as not real (modally distant). Consider the constructed examples (3) and (4).

1. (3) If [I had not met Mira’s father _P], [she would not be here _Q].

2. (4) If [I were a cat _P], [I would sleep all day _Q].

Example (3) considers a consequence (Q) of a hypothetical scenario (P) construed as counter to S’s reality. The speaker did meet Mira’s father. In example (4) an imaginary scenario (P) along with the possibilities that it would offer are entertained. In line with Dancygier and Sweetser, counterfactual is used here only for “specific interpretations involving a construed contradiction with [the speaker’s] ‘reality’” (2005, p. 58). Counterfactual conditionals about past events, as in (3), feature “both past-tense marking (the referent situation is in the past) and epistemic distance, hence past-perfect forms are used, with two morphological layers of marking” (2005, p. 57). Conditional if constructions with counterfactual meanings are not common in the data (4 instances). ^{Footnote 2} In (5), if I’d told my mom we could’ve saved him illustrates the instability of the speaker’s reality and the interaction and overlap between S’s reality and other non-realities (Figure 3).

1. (5) I kept getting visions ((gestures to head)) of eh like a tarantula or a spider ((gestures spider-like movements with hand)) crawling in biting my dad it being venomous and him dying and me feeling guilty ((gestures to chest)) because I was having these premonitions ((air quotation marks)) and I didn’t tell my mom and maybe if I’d told my mom we could’ve saved him (OCD 5)

In Figure 3 the experienced guilt is represented in R (S’s reality space) along with not telling mom about the premonitions. The dotted line for ‘tell’ in R indicates negation (didn’t tell). I kept getting visions sets up a space for the visions: R’. This space shifts S’s deictic origin through (scalar) translation along the m axis.

Fig. 3.

Simplified DSM for example (5).

From the perspective of S as an adult, R’ is distant on the m axis; the use of visions is epistemically distancing (visions are typically not real). In addition, air quotation marks in conjunction with premonitions indicate a negative epistemic stance. From the perspective of seven-year-old S, the epistemic status of R’ is uncertain. Are the premonitions real or not? The visions clearly have emotive consequences to S in R; seven-year-old S feels guilty and considers whether he should tell his mom about the spider biting his dad. The modal instability of R’ is represented in Figure 3 by a double-lined arrow along the m axis. Within the space for the visions (R’) at a point in S’s past (–t on the temporal axis) is a causative chain: spider crawling in, spider biting dad, spider venomous, dad dying. In the figure, this causative chain is represented in simplified form by three vectors. A compelling feature of R’ is the use of the nonfinite –ing forms to describe this chain of events; they construe the events and states of the visions as dynamic (e.g., it being venomous), and akin to the progressive form these nonfinite forms evoke “a sense of being closer to [S’s] personal experience” (Chilton, Reference Chilton2014, p. 129). At the same time, these forms provide little anchoring to the t and m axes, something that is in line with their construed status as epistemically uncertain and temporally unmoored. The use of because ties S’s experienced guilt to the events in R’ and thereby conflates R’ with R (construes the visions as real). Not telling S’s mom about the visions can only cause S’s dad to die if the events of the visions are real. Another space (R’’) is evoked by the admission maybe if I’d told my mom we could’ve saved him. R’’ is a modal reflection of R (i.e., across the m axis). The if-space is predicated on the visions being real; in R’’ (the if-space), S told his mom about the premonitions and his dad could be saved.

The OCCWG write that “the occurrence of an intrusive thought about a possible future harm is appraised by the person with OCD as indicating that he or she is responsible for preventing the harm” (1997, p. 672). Emotions like guilt and shame, and also hope, anticipation, and relief, rely on counterfactual reasoning (Byrne, Reference Byrne2007, p. 9). The guilt expressed in (5) depends on the construal of the scenario evoked by maybe if I’d told my mom we could’ve saved him, essentially “a comparison between how the event actually turned out and how it could have or should have turned out differently” (Byrne, Reference Byrne2007, p. 9). The emotion of guilt has been accounted for in the literature on OCD (Rachman, Reference Rachman1993; Shafran, Watkins, & Charman, Reference Shafran, Watkins and Charman1996; Tilghman-Osborne, Cole, & Felton, Reference Tilghman-Osborne, Cole and Felton2010) and features prominently in the data for this study.

7.2. predictive conditionals

The most common function of if constructions in the data is prediction (68 instances). Example (6) predicts (with some uncertainty as expressed by could) that if S does not wash her hands, a number of scenarios linked to contamination may occur.

1. (6) if [I don’t wash my hands _P] [I could do this _Q] [this could happen _Q] [I could pass it on to someone else _Q] [they could die _Q] [I could die _Q] (OCD 9)

Dancygier and Sweetser write that “prediction is a central sense of the conditional construction (2005, p. 42; also Dancygier, Reference Dancygier1998). It involves “setting up a hypothetical (typically future) mental space and attempting to predict its consequences based on knowledge of typical cause–effect chains and general world-knowledge” (Dancygier & Sweetser, Reference Dancygier and Sweetser2005, p. 122). Prediction is not only central to conditional constructions, but is essential to reasoning and decision-making in general. Dancygier and Sweetser argue that “[w]ithout engaging in prediction – in the construction of (and commitment to) some future scenarios as more likely than others – we could never make decisions or take action at all” (2005, p. 31). Because the testimonials primarily describe past experiences tied to OCD, most of the predictive conditionals in the data predict (future) consequences of past events, as in if I didn’t do it, someone in my family would die (OCD 5). In addition to describing past events, the past tense in conditionals can signal “epistemic distance” (2005), as in (constructed with future meaning) If he was found guilty, he would be sentenced to life in prison. In the data, however, the past tense is not used primarily to signal negative epistemic stance, but to describe conditional scenarios in the speaker’s past.

Normally, predictive conditionals “construct not just one single space involving P and Q, but also an alternative space involving ∼P and ∼Q” (Dancygier & Sweetser, Reference Dancygier and Sweetser2005, p. 35).

These alternative spaces are “incompatible” and “are both construed in context as subsequent causal and temporal developments from the same base or reference space, for the same portion of the timeline” (2005, p. 35). In connection to OCD, O’Connor et al. argue that “thinking of alternative scenarios can help us arrive at a sense of reality” but can also “increase doubt rather than certainty” and “[u]nder certain conditions people with OCD seem more susceptible to this doubting effect” (2005, p. 69).

Predictive if conditionals in the data foreground negative consequences. In (7), predictive if not P, then Q and if P then Q make salient negative outcomes: I was going to die and I would get the disease.

1. (7) I felt like if I didn’t look back ((looks over shoulder)) like six times or something ehm … that I was going to die, or ((sighs)) I felt like the number … like if I inhaled when I was looking at someone that … with like a disease then I would get the disease (OCD 2)

Dancygier and Sweetser write that “[t]he space directly mentioned in a conditional prediction is set up more saliently and directly than its implicit alternative” (2005, p. 36). Example (7) illustrates something significant to the OCD data: the space construed by S as undesirable/negative is foregrounded and explicit, whereas a desirable/positive space is evoked implicitly and is derivative of the negative space. Only in a few instances in the data are positive outcomes mentioned explicitly, e.g., if you complete these rituals to your like satisfaction that … it will ease the anxious feeling that you have (OCD 7).

The if-clause in (8) sets up a scenario, or “a model of some situation in the world” (Dancygier & Sweetser, Reference Dancygier and Sweetser2005, p. 29). In terms of DST, the if-clause evokes a deictic space and a consequence is predicted based on this space. In addition, a contrastive scenario (not doing anything wrong) along with its emotive consequence (not feeling guilt) is implicitly evoked.

1. (8) the guilt would eat at me if I did something that I felt was wrong (OCD 2)

How can construal of these alternative/contrastive spaces be accounted for? I suggest that this type of contrast can be represented as two reflected spaces in a DSM. Chilton (Reference Chilton2014) uses the transformation of reflection to account for true counterfactual meanings, as in: The guilt would have eaten at me if I had done something wrong [but I didn’t] (Section 7.1). The use of reflection to account for the evoked construal of (8) is an extension of this use of reflection. S’s reality space (R) corresponds to the implicit scenario of (8): ‘the guilt would not eat at me if I didn’t do something I felt was wrong’. In the extended context of (8) the speaker says that she never broke a single rule (example 9), so S’s actions are in line with the implicit scenario where S does nothing wrong (S’s reality space R). While not strictly counter to fact, the foregrounded if-space (R’) constitutes a reflection of R at the limiting point of the modal axis. The circumstances of R are reversed in R’. Figure 4 provides a DSM for example (8).

Fig. 4.

DSM for example (8) illustrating reflection across the modal axis.

Reflection bisects the m axis according to a boundary (line of reflection) and construes R and R’ as being in opposition (as incompatible alternatives). The dashed lines in the speaker’s reality space (R) indicate negation. ^{Footnote 3} The avoidance of R’ guides S’s actions and makes her follow every rule. Doing something wrong is represented as causing the guilt to eat at S (indicated by a force vector in the DSM).

Predictive conditionals in the data evoke causative construal. In (8) doing something wrong causes the guilt. Causative meaning can also be cued by the lexical items because and so, as in (9).

1. (9) I was always the kid to follow the rules perfectly because if I didn’t I would feel this really bad shame and guilt and so I never broke a single rule (OCD 2)

Similar causative meaning is also evoked in the data by because and or in combination with have to, as in I have to do X because if I don’t Y will happen and I have to do X or Y will happen.

7.3. exceptive conditionals

Similar to predictive if conditionals, unless constructions can be used to foreground negative scenarios that may come to fruition if compulsions are not carried out (Dancygier, Reference Dancygier2002; Dancygier & Sweetser, Reference Dancygier and Sweetser2005, ch. 7; Declerck & Reed, 2001, ch. 13). Unless conditionals are less common in the data (8 instances) than if conditionals.

1. (10) for me my thought was you’re going to be sick or your friend’s going to fall out with you … UNLESS you do this unless you text your friend five times unless you … go unless you stop walking on all the pa- the cracks in the pavement unless you … ehm choose different cutlery to eat your food with unless you eh walk back and forth five times (OCD 11)

In (10), two scenarios (Qs) are evoked by the expression you’re going to be sick or your friend’s going to fall out with you. These scenarios are foregrounded by S (explicit and construed as the default). The subsequent unless-clauses (Ps) cue a number of ‘exceptive’ scenarios (Dancygier & Sweetser, Reference Dancygier and Sweetser2005) to yield the meaning: Q; [(not Q) if P] (Dancygier, Reference Dancygier2002; Dancygier & Sweetser, Reference Dancygier and Sweetser2005, p. 184). Example (10) “states Q as a basic default situation and then adds the If P, ∼Q scenario as an exceptional alternative” (Dancygier & Sweetser, Reference Dancygier and Sweetser2005, p. 187). The unless construction in (10) allows S to foreground two negative scenarios that will happen unless exceptive measures are taken. The use of you in (10) instead of I introduces a degree of distance between S and the thoughts and potentially indicates that the thoughts are construed as having a source external to S, a type of disassociation between OCD and the self (Fennell & Liberato, Reference Fennell and Liberato2007, pp. 322–323).

Much like if constructions, unless constructions evoke contrastive meaning (alternativity) between “mutually exclusive scenarios” (Dancygier & Sweetser, Reference Dancygier and Sweetser2005, p. 187). In the data, as we have seen for previous examples, scenarios considered immoral, scary, or unreal are explicit, whereas their alternatives are implicit. The default state for the speaker of (11) is discomfort, and obsessive worry is described as the only way around this sensation.

1. (11) I don’t feel comfortable unless I’ve got something to obsess about which sounds ridiculous because like … the only way for ME to get rid of a particular worry is to then find something else to worry about so I will … sit and wait look for more dangers (OCD 9)

In (11), unless is used to convey that only worrying about more dangers can alleviate the discomfort of worrying, which is described as S’s default experience.

7.4. reduced-q conditionals

Doubt and catastrophizing are features of OCD that can be expressed through what-if constructions (Abramowitz, Reference Abramowitz2006; Nikodijevic et al., Reference Nikodijevic, Moulding, Anglim, Aardema and Nedeljkovic2015; OCCWG, 1997; O’Connor et al., Reference O’Connor, Aardema and Pélissier2005; Shafran, Reference Shafran, Abramowitz and Houts2005; Van Schalkwyk et al., Reference Van Schalkwyk, Bhalla, Griepp, Kelmendi, Davidson and Pittenger2016). Doubt and catastrophizing are related; doubt can concern the past, the present, or the future (What if I did … ? What if I am … ? What if I will do … ?) and catastrophizing generally concerns the relevance of the past/present to future events: “Catastrophizing is the tendency of individuals to apply a ‘what if…?’ questioning style to potential problematic features of their life and, in worriers, this usually results in the consideration of worse and worse outcomes” (Shafran, Reference Shafran, Abramowitz and Houts2005, p. 237). The use of what if evokes a new deictic space through transformation of S’s base reality space. A what-if space can be constructed through translation (scalar repositioning on the modal axis) for meanings involving an uncertain future, as in (12), in which the use of maybe reinforces a scalar construal.

1. (12) oh I’ve trod in dog poo ah that means like maybe next year I’ll go blind and maybe the year after that I’ll want to kill myself … oh my god and what if (OCD 1)

Uses of what if can also evoke deictic spaces through reflection across the modal axis for contrastive meaning, as in (13), which evokes doubt (what if I AM crazy) and catastrophizing (what if I go crazy and this happens).

1. (13) everything is a what if everything becomes for me what if I AM crazy what if I am … you know… what if I DO want to hurt these people what if I DO want to do whatever what if I go crazy and this happens (OCD 3)

Emphasis is placed on the words AM and DO in a manner that evokes their opposites as the default. They cue construction of contrastive spaces, as in Figure 5. The use of what if I AM crazy cues the conception of ‘I am not crazy’ in R (dashed line), along with its modal reflection ‘I am crazy’ in R’ (solid line) through a comparative construal that balances on a modal mid-boundary. Being crazy (and then going crazy) causes something to happen, an event represented in R’ (Figure 5) in simplified form as a vector at a time t _i in the future of R’. A number of content dimensions are aligned with the modal axis for contrastive construal: control (‘control’ vs. ‘no control’), mental state (‘crazy’ vs. ‘not crazy’), desire to hurt people (‘want to’ vs. ‘do not want to’).

Fig. 5.

Partial DSM for example (13) what if I am crazy … and this happens.

Unlike predictive if constructions, what-if constructions leave Q implicit. Declerck and Reed refer to what-if constructions as “reduced-Q conditionals” (2001, pp. 390–391) and write that in what-if constructions “what can represent various Q-clauses (What shall we do?, How will we react?, etc.)” and can “be short for something like ‘what are the implications?’” (2001, p. 391). They also suggest that what if can be used to consider an “alternative” to some idea as in “… What if the murder was committed with a different knife” (2001, p. 391). In (14), S realizes that she checked on her rabbits only five minutes ago and that most likely nothing has changed since then, but the evoked alternative scenarios (what if they just died what if they just did this or what if they’re sick) make her initiate her checking routine again.

1. (14) if I hear a noise or I hear something like I don’t know the rabbit like ran or jumped or something I start obsessing what if they just died what if they just did this or what if they’re sick and even though I just checked on them five minutes ago I have to get out of bed go downstairs and do the whole thing over again (OCD 6)

What-if constructions illustrate “intolerance to uncertainty” (OCCWG, 1997), a feature of OCD that is closely linked to doubt (Section 2.1).

7.5. antonyms

Conditional constructions are used in the literature on OCD to exemplify patterns of reasoning and to suggest screening and assessment. Antonyms, on the other hand, are much less common in such accounts, if they are considered at all. This is noteworthy because antonymous uses of lexical items reinforce and specify the (contentful) nature of the construals evoked by conditional constructions. In accordance with LOC, antonymy is regarded here as “a conceptual relation” of contrast (Jones et al., Reference Jones, Murphy, Paradis and Willners2012, p. 127). Through antonymous uses of lexical items an “underlying contentful meaning structure is divided into two parts” (2012, p. 134) that are construed as being in opposition. In the OCD data, antonymous lexical items evoke contrast according to a range of content dimensions. Modal, moral, emotive, and perceptual dimensions are prominent, as are dimensions tied to actions and control.

1. (15) it would be like looking back on past past incidents where they weren’t really anything but my brain would now layer stories overtop and say you did this despicable thing to this person you hurt this person when in reality none of that’s true but my brain would force me to buy into it and and it was really nasty and I’d be ruminating about that for days because you know I don’t I don’t want to be a bad person ((points to chest with both hands)) I don’t think I’m a bad person and my brain was trying to TELL me I was a bad person and anyone with OCD knows … how kind of painful that is and hurtful and scary because it’s completely opposite of what you want it’s ego-dystonic which is why OCD is so painful (OCD 3)

Example (15) contrasts what is subjectively real (R) to what is experienced as a forced counterfactual reality (R’) through expressions like layer stories overtop and in reality none of that’s true. R’ is explicit but (15) also evokes R, a space in which S did nothing despicable to hurt another person. These two spaces (R and R’) are reflections across the modal axis: what is real in R is not real in R’ and vice versa (Figure 6). A number of lexical items establish and reinforce this contrast. The use of a bad person, especially in the context of negation (don’t think I’m, don’t want to be) reinforces the modal contrast (real vs. not real) by aligning it with a moral contrast and a contrast for desirability. The modal, moral, and desirability dimensions are also integrated with an emotive one through uses of painful, hurtful, and scary to create a joint contrastive construal through integration of these quality dimensions. The contrastive boundary construal is further reinforced by the use of the adjectives opposite and ego-dystonic.

Fig. 6.

Integrated dimensions for contrastive construal in example (15).

The ‘do not want–scary–bad–false’ scenario of R’ is explicit and foregrounded by S. The ‘want–calming–good–true’ scenario of R is only represented implicitly and mainly through negation (I don’t want to be a bad person).

Contrast can also evoke balance through force-dynamic construal. In the data, compulsions are described in terms of symmetry and counterforce (Rachman, Reference Rachman1993; Salkovskis Reference Salkovskis1989; Taylor, Reference Taylor, Frost and Steketee2002). Alongside force, balance relies on boundary construal and oppositeness. In (16) ‘hitting my left hand’ establishes sensory balance.

1. (16) I did have weird compulsions too like I would have to … if I like hit my hand on something my right hand I would have to hit my left hand on it so it felt the same way (OCD 2)

Balance is reinforced in (16) through the use of antonymous right and left and through the description of a resulting sensation: so it felt the same way. The use of so cues a causative relation between action and result. In example (17) from the same testimonial, a number of dimensions are integrated for contrastive construal and antonymous uses of lexical items establish and reinforce contrast.

1. (17) some of my intrusive thoughts were … ehm … I’m just going to be straight up with you guys … were around like inappropriate images and things that … ehm … I felt were really wrong for me to be thinking about at a young age not only were they inappropriate in like taboo ways but they were also sometimes really scary thoughts like I would think things like my family … like I would over and over in my head something would play like … ehm … like someone I love would get shot but then in my head to make myself feel better I would like imagine myself taking the bullet for that person so that I wouldn’t feel guilty for picturing someone I love getting shot … you guys it sounds insane and it kind of is (OCD 2)

The use of inappropriate, wrong, and taboo establish contrast for morality. Note that, as for (16), their opposites (appropriate, right, and acceptable) are implicit and not foregrounded. The moral contrast is aligned with an emotive one through scary.

Two imagined actions are balanced: picturing someone I love getting shot and imagining myself taking the bullet for that person. The former action is intrusive / not controlled, whereas the latter is controlled and aimed at establishing balance relative to a moral/emotive dimension introduced by guilty. These dimensions are integrated and construed as incompatible alternatives according to a single boundary (Figure 7). A force dynamic is introduced through the action of imagining myself taking the bullet for that person; causation is expressed by so that I wouldn’t feel guilty and to make myself feel better, which present desired results. Unlike (15), example (17) does not explicitly evoke a modal contrast (real/unreal) through words like reality or true but, as in (15), emotions like fear and guilt are construed as proximal to S. The intrusions in (17) are emotively ‘real’ and yet the speaker suggests that her experience is ‘not real’ when she says that it sounds insane and it kind of is. A double-lined arrow in Figure 7 indicates that S’s deictic origin is in flux and shifts between R and R’. Similar to the expression it feels so real [but I know it is not], emotive proximity is contrasted with modal proximity (Section 6). The intrusions are modally distant (insane) but emotively proximal (scary and guilt-inducing).

Fig. 7.

Integrated dimensions for contrastive construal in example (17).

Research on OCD has accounted for the neutralizing function of compulsions (Salkovskis, Reference Salkovskis1989). Taylor writes that “[c]ompulsions can be overt (e.g., cleaning) or covert (e.g., thinking a ‘good’ thought to undo or replace a ‘bad’ thought)” (2002, p. 1). In (17), the thought of taking a bullet for a loved one serves to ‘undo’ the scary intrusive image and achieving balance is predicated on a relevant contrast according to morality along with an action (the thought) with potential for influencing this contrast. In (16), the balancing action is hitting my left hand and balance is tied to sensory perception (hitting my left foot means I have to also hit my right foot) (Hartman, Reference Hartman2017, for force construal and Tourette Syndrome). In (18), two thoughts are construed as opposites: oh my God you’re gonna stab her and you’re so lucky to have her. The action of balancing these thoughts is described as ‘rhyming’, and balance is sought according to morality. Similar to previous examples, (18) evokes two spaces reflected across the modal axis.

1. (18) I developed a weird compulsion ((makes circles with hands toward body)) where I would rhyme so like I would get scary words or phrases in my head constantly … and I’d have to rhyme it with a POSITIVE thought so like for example if I was cutting something up ((gestures cutting)) and I was thinking oh my God you’re gonna stab her oh my God you’re gonna stab her I would have to rhyme it with … you’re so lucky to have her oh my gosh you’re so lucky to- like and I would just DO it ((makes fast circles with hands)) and it became … very automatic to me (OCD 12)

Uses of, for instance, the adjectives scary, positive, and lucky reinforce the construed contrast. The admission I’d have to rhyme it indicates that the positive thought (you’re so lucky to have her) neutralizes or cancels out the negative thought (oh my God you’re gonna stab her).

Antonymous uses of lexical items as in (15)–(18) illustrate the speakers’ preoccupation with, and linguistic–conceptual foregrounding of, only one end of a number of integrated meaning structures. Insight into the kinds of construal reflected in (15)–(18) can be valuable in responses to and treatment of OCD. Contrastive (boundary) construal constitutes a mode of configuring conceptual content that leaves little room for inferences and actions of a gradual or incremental nature. Also, the integrated nature of the dimensions potentially influences the conceptualizing person’s capacity to single out and reconsider any one of these dimensions in isolation. For instance, targeting an action, such as a compulsion, for reconsideration and/or change may require also adequately addressing any links between the action and interrelated dimensions, e.g., moral, emotive ones (see Murphy & Perera-Delcourt, Reference Murphy and Perera-Delcourt2014, for the relation between a qualitative approach and treatment of OCD). The mutability and multidimensional character of the speakers’ conceptions of reality indicate a need for careful consideration of diverse types of experiential proximity to the conceptualizer above and beyond modality (e.g., emotion, morality, control).