The Structure Of Misunderstandings

The structure of misunderstandings* Izchak M. Schlesinger and Sharon Hurvitz Hebrew University of Jerusalem

In this paper we introduce a detailed and multi-faceted characterization of misunderstandings. The proposal attempts to capture the structure of misunderstandings in terms of several constructs: the message as intended by the speaker, the message as construed by the hearer, and the message as understood by an ‘objective’ judge. In addition, we suggest that the message the speaker intends the hearer to retrieve and the hearer’s perception of the speaker’s intentions should also be taken into account. Misunderstandings can also be classified according to the phase of the comprehension process at which they occur (the perception of the speaker’s message, its linguistic processing, discovering the implicatures, and so on). Keywords: conversation, discourse, misunderstanding, non-understanding, speaker intention

1. Introduction Behind nearly all verbal interactions there lurks the danger of misunderstanding (cf. Blum-Kulka and Weizman 1988), and the allegation of pretending to misunderstand is common in controversies (Dascal 2003: 280–292). In the present paper we attempt a conceptual clarification of misunderstandings, real and pretended. The phenomenon of misunderstanding has been discussed in the literature from a variety of different angles and has received special attention in sociolinguistic and discourse analyses. Differing approaches to misunderstanding have chosen to focus on distinct aspects of the phenomenon, a fact which has led to varying characterizations, relying on differing methodological frameworks. Some researchers have attempted to characterize at what stage in the conversation a misunderstanding is detected and corrected (cf. for example, Bazzanella and Damiano 1999), while others have concentrated on how misunderstandings are managed in discourse (cf. for instance, Humphreys-Jones 1986; Schegloff 1987, 1992; Trognon and Saint-Dizier 1999; Bazzanella and Damiano 1999; Dascal 2003: 313–316). Pragmatics & Cognition 16:3 (2008), 568–585.  doi 10.1075/p&c.16.3.07sch issn 0929–0907 / e-issn 1569–9943 © John Benjamins Publishing Company



The structure of misunderstandings 569

Grimshaw (1980) discusses responses to and other effects of misunderstandings, and Taylor (1986) has broached the question of how we know that the hearer has understood the speaker. Yet others (Bazzanella and Damiano 1999; Weigand 1999, for example) have proposed a typology of misunderstandings according to their causes. Dascal (2003: 280–302) provides a useful overview of major studies carried out with respect to misunderstanding. The purpose of the present paper is different. We attempt to capture the structure of misunderstanding in terms of the assumptions and intentions of the interlocutors as mediated by their utterances. In addition to the standpoints of the speaker and the hearer, we consider a ‘neutral’ standpoint. The proposed parameters, considered jointly, permit to identify types of misunderstandings that previous analyses have not recognized as such. An additional classificatory principle, based on the various aspects of the comprehension process at which misunderstandings can occur, is developed in the last section of the paper. The term ‘misunderstanding’ will be used here in a wider sense than that in common parlance: any departure from “perfect” communication we will call ‘misunderstanding’, which includes what commonly are called ‘partial understanding’ and ‘non-understanding’ (see Section 8). 2. Hearer-caused and speaker-caused misunderstandings Understanding, according to Locke (1960: III, 9, 6), occurs when words “excite in the Hearer exactly the same Idea they stand for in the mind of the Speaker”, and this view is shared by many classical authorities from Aristotle to Hobbes (Taylor 1986: 82). A more modern approach refers to “the correspondence between A’s and B’s [i.e., the speaker’s and the hearer’s] perception of a ‘conversational demand’” (Dascal 2003: 306).We can capture this approach by letting M(S) stand for the message that the speaker (S) intends her utterance to convey, and M(H) for the message that the hearer (H) ascribes to the utterance — where the term ‘message’ remains unanalyzed for the time being (but see Section 9) — and defining understanding as M(S) = M(H). On the face of it, a misunderstanding might accordingly be symbolized simply by M(S) ≠ M(H). This formula, however, acknowledges only the fact that a misunderstanding has occurred; it says nothing about the type of misunderstanding — whether it is due to the speaker’s inappropriate phrasing or to the hearer’s shortcoming, whether it has been unintended or faked, and so on. As will be shown, additional concepts are needed for a taxonomy of misunderstandings.

570 Izchak M. Schlesinger and Sharon Hurvitz

Consider, first, an important distinction that lies behind the following two misunderstandings: (1) A1: Can you give me a ride home? B: Oh, I’m afraid I have a flat tire. A2: But can you take me home?

A obviously fails to retrieve what is implicated by B’s utterance, namely that he cannot take her home because of the flat tire. The misunderstanding must be laid at the hearer’s (i.e., A2’s) door: hearers can usually be expected to understand what B has here implicated. Compare this to (2) S1: Suddenly the rain started to pour down, and he took cover in a house. H: Did he know the people there? S2: What do you mean? Wasn’t he living with them?

By “a house” S1 meant ‘his own house’, whereas H understands this as ‘someone else’s house’. The speaker could have taken “special care to prevent misunderstanding” (Dascal 2003: 315), but in the present example he used a misleading expression in referring to his own house and thus caused this misunderstanding.1 The simplistic formula M(S) ≠ M(H) does not capture the distinction between misunderstandings caused by the speaker — as in (2) — and those caused by the hearer — as in (1). What is missing here is a link mediating between M(S) and M(H). The speaker’s and hearer’s minds after all do not communicate with each other directly, but only (in the cases to be considered in this paper) through the medium of a linguistic utterance. One can abstract from the interpretations of the particular speaker and hearer, and the message carried by an utterance can be understood on the basis of the conventions of language and the context of the utterance. We suggest, then, that in addition to M(S) and M(H) a third concept be introduced into our conceptual apparatus: the way a “faultless hearer” (Hockett 1986: 53, 57) understands the speaker’s utterance.2 Let us symbolize this as M(A) where A stands for the analyst (assumed to be an idealized, faultless hearer). The analyst’s interpretation, by definition, adheres to the semantic and pragmatic rules of the language,3 whereas the speaker and/or the hearer may fail at times to do so, and this failure may result in speaker-caused or hearer-caused misunderstandings.4 In (1), the analyst will understand the message in the way intended by the speaker — M(S) = M(A) — whereas the way the hearer understands it differs from both M(S) and M(A). In (2), the analyst holds that the utterance conveys a message that differs from the one the speaker intended it to convey. The symbol N will be used for a message that differs from M in the example analyzed; that is, besides M(A) and M(H) there may be N(A) and N(H). Examples (1) and (2) will accordingly be symbolized as in [1.1] and [1.2], respectively:



The structure of misunderstandings 571

[1.1] M(S) M(A) N(H) [1.2] M(S) N(A) N(H)

The difference between M and N may lie in either (a) N differing from M or (b) N not including an aspect of M (this point will be taken up in Section 8). A misunderstanding may be both speaker-caused and hearer-caused: the speaker may phrase her utterance in a way that does not correspond with her intentions and the hearer erroneously understands this utterance in a way that does not correspond with the way the analyst reads it. Using the symbols N1 and N2 for two messages that each differ from M. we can symbolize this as in [1.3]: [1.3] M(S) N1(A) N2(H)

Such misunderstandings are probably rare.5 Consider now that an utterance may convey a message that is ambiguous in its particular context; for instance:6 (3) H1: How was your vacation? Go anywhere interesting? S1: I couldn’t afford to, and even if I had the money, I had a lot of marking to catch up on. Then I was interviewing all last week. H2: Yes, of course. S2: What about you? H3: Oh, well I don’t do admissions any more. Used to, of course. S3: No, I mean did you go anywhere interesting? (after Falkner, 1997: 118)

What S2 meant becomes apparent only by his reaction in S3. The question “What about you?” (S2) might be understood in either one of two ways, both ‘sanctioned’ by the analyst: as referring to the work load (proximal topic) or to the vacation (distal topic). We therefore deploy M1 and M27 — rather than M and N — and a slash for disjunction: [1.4] M1(S) M1/ M2(A) M2(H)

This, too, is a speaker-caused misunderstanding, since it is due to the speaker having failed to choose an unambiguous way of expressing herself in a way that does not leave room for misunderstandings. As a result, the ‘analyst’ construes her utterance as M1/M2. Ambiguity is also involved in double entendres. In [1.5], the speaker intends to communicate two meanings, M1 and M2, and M1& M2(A) indicates that her utterance was formed according to her intention. Since H ‘gets’ only one of the two meanings (namely, M1) and is deaf to the other one, this is a misunderstanding (of a perhaps rather harmless kind): [1.5] M1& M2(S) M1& M2(A) M1(H)

572 Izchak M. Schlesinger and Sharon Hurvitz

This is a hearer-caused misunderstanding (albeit only a partial one, since H at least retrieves one of the messages, M1). The foregoing include some of the more common types of misunderstanding. The list, however, is not exhaustive. There are still other possibilities. For instance, a hearer may detect an unintended ‘double entendre’; this would be symbolized by M1(S) M1&M2(A) M1&M2(H). Instead of enumerating all possibilities, we will deal in the following sections with types that require an extension of the conceptual apparatus introduced in the foregoing. 3. Further types There are types of misunderstanding that differ in important ways from those dealt with in the previous section. In everyday parlance, some of these will perhaps be called partial understanding, or, perhaps, even ‘full’ understanding, but it is proposed here to use ‘misunderstanding’ as a theoretical term; we return to this question in Section 6, below. It turns out that a formal description of these types cannot be carried out by means of the concepts introduced so far, and our conceptual apparatus will have to be augmented. Consider the following cases, a–h: (a) A speaker is uncooperative and phrases her utterance with the intention that the hearer misunderstands it; and the hearer in fact misunderstands it. (b) As in (a), but counter to the speaker’s intention, the hearer does understand her. (c) The speaker’s utterance involves a double entendre, but she does not want the hearer to understand the secondary meaning. The hearer in fact does not understand it. (d) As in (c), but the hearer does in fact understand it. (e) S expresses herself infelicitously, in a way that will usually engender a misunderstanding, but H nevertheless understands what she intended to say. For instance, H may know S well and intuitively detect the latter’s imperfectly expressed intention. (f) The hearer detects a second meaning that was not intended by S. (g) The hearer only pretends to have misunderstood the speaker. For example: (4) Client: Waiter, what’s this fly doing in my soup? Waiter: It looks like the breaststroke, sir.



The waiter only pretends to understand the client’s words as an informative question but in fact knows that the client wanted to complain.

The structure of misunderstandings 573



(h) The speaker contrives a misunderstanding, that is, she manages to make H understand her in a way that S then pretends not to have intended. For example: (5)



S1: Do you know what is nice about you? H: No. S2: Too bad, I had hoped that you would be able to tell me. (after Falkner 1997: 112)

Unlike the speaker in (4), the speaker here presents a different analysis of her own previous utterance. In the following sections we introduce additional concepts to deal with these, and several other cases.

4. Speaker’s intention The term ‘speaker’s intention’, as used in the present paper, stands in need of clarification. M(S) has been defined as the message that the speaker intends her utterance to convey. Normally, this is also the message the speaker intends the hearer to understand, but not always. Occasionally a speaker may phrase her utterance with the intention that the hearer misunderstands it. Consider: (6) Mother: Have you done [a] the washing up and [b] put everything away? Little girl [=S]: I have done the washing up. M(S) = I have done [a]; (implicated:) but not [b].

Suppose now that while the little girl has truthfully refrained from denying that she hasn’t fully complied with her duties, she hopes that her mother won’t notice what is implicated by her answer. While she intended her utterance to convey the full message, including the implicature (‘I have not done [b]’), she intends her mother to misunderstand the message i.e., to retrieve M(S) without the implicature. Suppose, further, that her (perhaps momentarily inattentive) mother retrieves only what the girl intended her to: ‘I have done [a]’, which differs of course from the full M(S). This case, then, might be analyzed: M(S) M(A) N(H), where N(H) equals M(S) minus the implicature, ‘I have not done [b]’. Now, this formula does not show what the girl intends her mother to understand. What is needed here is an additional concept of speaker intention alongside that of M(S). We propose to use the term speaker’s intention — or S-int., for short — for the message the speaker intends the hearer to retrieve,8 and to retain M(S) — or M1/M2(S), M1&M2(S) — for the message the speaker intends the utterance to convey. Usually, the two kinds of intention will be identical, but in our example, the two will differ. The formalization for that example will be:

574 Izchak M. Schlesinger and Sharon Hurvitz

[3.1] M(S) M(A) N(H)

S-int.: N(H)

Such divergences between M(S) and S-int. are by no means rare. One has only to think of a witness testifying at court, who bewares of uttering a lie but hopes that the statement implicated by her answer will escape notice. Or think of someone telling the truth in technical language which she knows the hearer to be unlikely to understand. Opinions may be divided as to whether an example like this deserves the appellation ‘misunderstanding’ (because the mother failed in retrieving the implication contained in M) or not (because she understood the utterance in the way the girl wanted her to). But, as stated, we propose to use ‘misunderstanding’ as a theoretical term which stands for a concept that may differ in several respects from the pre-theoretical concept. There are many other cases that are not usually called misunderstanding, which we will discuss presently. The child’s intention in (6) may be frustrated: her mother may catch on and retrieve the implicated meaning. In that case we write: [3.2] M(S) M(A) M(H)

S-int.: N(H)

The distinction between the two kinds of intention, M(S) and S-int., will also be needed in analyzing a double entendre which the speaker intends the hearer not to notice (say, in order to make fun of him or show off in front of a third party). Then there are two possibilities: 1. The addressee in fact fails to notice the secondary meaning: [3.3] M1&M2(S) M1&M2(A) M1(H)

S-int.: M1(H)

2. The addressee “disappoints” the speaker and does recover the secondary meaning: [3.4] M1&M2(S) M1&M2(A) M1&M2(H)

S-int.: M1(H)

The S-intention is what distinguishes [3.3] from the misunderstood double entendre analyzed in [1.5]. In [1.5], it might also have been appropriate to add S-int., that is, to adopt [3.5] instead of [1.5]. [3.5] M1&M2(S) M1&M2(A) M1(H)

S-int.: M1&M2(H)

But since the S-intention, M1&M2(H), does not diverge from the speaker’s intentions –M1&M2(S) — it can be dispensed with or else optionally added. The same holds for the other examples in Section 2. A speaker may be unaware of the fact that what she says has two meanings (i.e., a kind of unintended double entendre). For instance:



The structure of misunderstandings 575

(7) Finance minister: We have been standing on an economic precipice, and we have taken a great step forward.

Presumably the minister meant both clauses to be taken metaphorically (we have been saved from a calamity by the steps we have taken) and did not intend to convey the meaning that some of his audience will gleefully note, namely, that ‘we’ have fallen into the abyss. But the audiences presumably know that this was not the speaker’s intention, which is why the joke is on the minister. The analysis will be: [3.6] M1(S) M1&M2(A) M1&M2(H)

S-int.: M1(H)

5. Hearer’s perspective We now introduce an additional concept that is required for the analysis of yet other cases. Suppose the mother in example (6) correctly retrieves the implicature in M(S) and also recognizes the girl’s intention to have her mother misunderstand her. Then we will want to show in the analysis not only the speaker’s intention as to what the hearer should retrieve but also the hearer’s perception of the speaker’s intention. We call this H-perspective and insert it in the second row of the formula. H-perspective, or H-persp., for short, relates to (a) what H believes S to have intended her utterance to convey, and (b) what H believes to be the S-intention: [4.1] M(S) M(A) M(H) S-int.: N(H) H-persp.: M(S) N(H)

This formula shows that while the hearer correctly understood the full message with its implicature — M(H) — she also realizes (a) what the speaker intended the utterance to mean, namely M(S), and (b) that the speaker intended her to misunderstand it: N(H). In this example, H-perspective is identical to the corresponding M(S) and S-int. But the H-perspective may be mistaken. Imagine a scenario where the mother wrongly suspects the little girl of intending that the implicature ‘I did not put everything away’ go unnoticed; the girl, though, was completely guileless: her S-intention was M(H). Then, instead of [4.1], we have: [4.2] M(S) M(A) M(H) S-int.: M(H) H-persp.: M(S) N(H)

There are still other possibilities. When S has expressed herself infelicitously, H may nevertheless understand what she intended to say. For instance, H may know S well and intuitively detect the latter’s imperfectly expressed intention. If H is an expert, say, a medical doctor, he may understand imperfectly described symptoms.

576 Izchak M. Schlesinger and Sharon Hurvitz

Or H may know that S is a native speaker of another language and ‘back translate’ the utterance into that language. Or the speaker may intentionally use a devious way of expression, expecting the hearer — but perhaps not anyone else — to guess at the intended message; which he does. In short, H may ‘bypass’ the infelicitous phrasing of S: [4.3] M(S) N(A) N(H) S-int.: M(H) H-persp.: M(S) M(H)

Such ‘bypassing’ may be detected in subsequent turns; e.g., when H corrects the first speaker (by repeating her words with a correction or by saying: “you mean…”, etc.). But a hearer’s attempt at bypassing can be erroneous: the speaker may have meant what she said (that is, what the analyst and the hearer believes her utterance to convey). As in [4.2], this will result in a divergence of H-persp. from the corresponding symbols in the first line, but the H-persp. will differ from that in [4.2]: [4.4] M(S) M(A) M(H) S-int.: M(H) H-persp.: N(S) N(H)

While the hearer correctly construes what the speaker said (this is indicated by M(H) in the first line), his lack of ‘trust’ in the felicitously worded utterance leads to his misunderstanding of what the speaker meant. Now, the speaker may not only have the intention that the hearer understand her utterance in a certain way but also intend the hearer to have a certain perspective. In [4.4], for instance, she may intend the hearer to have the H.-persp. he actually entertains — N(S) … N(H) — or else a different perspective. In other words, there can be a second-order S-intention, the object of which is the hearer’s perspective. Further, the latter can in its turn be the object of a second-order H-perspective. S-int. and H-persp., then, can alternate and apply recursively. We will not illustrate here the symbolization of such higher-order S-int. and H-persp., which in any case are much rarer. 6. Understanding and misunderstanding defined We can now define understanding and misunderstanding in terms of the concepts introduced so far. Consider first understanding. The prototypical example of successful communication would be: [5.1] M(S) M(A) M(H) S-int.: M(H) H-persp.: M(S) M(H)



The structure of misunderstandings 577

When an utterance is ambiguous and H understands it in the way intended by S, this may be symbolized as follows (compare this to [1.4]): [5.2] M1(S) M1/ M2(A) M1(H) S-int.: M1(H) H-persp.: M1(S) M1(H)

Again, when H understands an utterance as a double entendre, we write: [5.3] M1&M2(S) M1&M2(A) M1&M2(H) S-int.: M1&M2(H) H-persp.: M1&M2(S) M1&M2(H)

There are three properties that [5.1]-[5.3] have in common, and we suggest that these be regarded as necessary and jointly sufficient features of ‘understanding’: 1. The message recovered by the hearer is identical to the one the speaker intended the utterance to convey; in symbols: X(S) = X(H), where X stands for any expression (e.g., M, M1&M2). 2. The S-int. does not differ from what the hearer actually recovered, namely X(H). 3. There is no discrepancy between the H-persp. and the corresponding X(S) and S-int. As for misunderstandings, it is proposed that this term be used to refer to any communicative interaction that falls short of meeting any one of the three above conditions defined in the preceding. ‘Misunderstanding’, in the proposed terminology, thus applies to cases like (a)–(f) in Section 3 — which have been analyzed in Sections 4 and 5, as well as what is sometimes called ‘partial understanding’ and ‘non-understanding’ (the latter concept will be discussed in Section 8 below), in short, to any divergence from ‘perfect’ understanding, as defined in the preceding. We thus deploy ‘misunderstanding’ as a theoretical term, whose usage departs in some respects from that of the corresponding pre-theoretical one. Next we turn to cases (g) and (h) considered in Section 3, which necessitate the introduction of additional symbols. 7. Pretended misunderstandings Not all failures of communication are genuine mistakes. Misunderstandings may be feigned or faked. When the hearer feigns a misunderstanding, this will become evident in his response to the speaker. In a faked misunderstanding, the speaker attempts to mislead the hearer by contriving (usually with the intention of being funny or sarcastic) what may look like a misunderstanding. This will become apparent when the speaker subsequently responds to the hearer’s turn.

578 Izchak M. Schlesinger and Sharon Hurvitz

(a) Feigned misunderstandings. An example of a feigned misunderstanding is (4), given in Section 3. Here is another one: (8) Eileen: I have really missed Paddy. Herb: Yes, me too — by two inches.

(after Nerlich and Clark 2001)

Note, first, that here the three conditions of correct understanding are clearly satisfied: Herb presumably knows what Eileen meant by “missed”. In (4), the waiter, unless he is uncharacteristically dumb, knows that the guest wants to complain. The formula applicable to the first speaker should therefore be the same as that of a correct understanding; cf. [5.1]. What makes these examples different from [5.1] is that implicit in the second speaker’s (the waiter’s, Herb’s) turn is an interpretation of the previous speaker’s turn. This implicit interpretation diverges from the one which we, as analysts, impose on the first turn, and will be represented in a separate line in lower case letters: Herb pretends to understand Eileen’s words as conveying m2, while they ‘really’ meant, and were understood, as M1: [6.1] M1(S) M1(A) M1(H) S-int.: M1(H) H-persp.: M1(S) M1(H) m2(s) m2(a) m2(h) S-int.: m2(h)

(b) Faked misunderstandings. These differ from feigned misunderstandings in that S manages to make H understand her in a way S then pretends not to have intended. Examples are (5), repeated here, and (9). (5) S1: Do you know what is nice about you? H: No. S2: Too bad, I had hoped that you would be able to tell me. (9) [Alice and Bill are discussing the earthquake damage in their apartment] Alice1: You know, what I regret most is the crimson vase we got for a wedding present. Bill: Oh, I didn’t notice that this got broken, too. Alice2: It didn’t, and that’s what I regret.

Unlike the above examples of feigned misunderstanding, the speaker here presents a different analysis of her own previous utterances: Alice2 construes her own previous utterance, Alice1, in a way it has not been understood before. Similarly, in (5) S2 makes it appear that S1 should be understood in a way that differs from what the analyst, and most probably H, too, understood.



The structure of misunderstandings 579

The following formalization pertains to the first turn (i.e., S1 in (5) and Alice1 in (9)). The first line shows what actually occurred, and the line in lower case — what S2 and Alice2 pretend to have occurred. [6.2] M1&M2(S) M1(A) M1(H) S-int.: M1(H) H-persp.: M1(S) M1(H) m2(s) m2(a) m2(h) S-int.: m2(h)

Such faked misunderstandings lie at the root of a large proportion of humorous interchanges. They do not even require any interchange: a single turn will do. In (10), for instance, the effect is achieved by the first part of the utterance being understood in a way belied in the second part: (10) [At a friend’s wedding:] I, too, did not know what is true happiness till I married, and then it was too late.

8. ‘Non-understanding’ In the literature, a distinction is made between misunderstanding and non-understanding. It should be appreciated, however, that this is not a simple dichotomy. Total absence of understanding is relatively rare; partial understanding seems to be more usual. It makes perfect sense to state, for instance: “I understand what you say, but I don’t know what you mean by it”. In actual cases of ‘misunderstanding’ there are only some aspects of the message that the hearer fails to recover completely, while others may be transmitted intact. In (1), for instance, the hearer presumably understands that the speaker has a flat tire; what he fails to retrieve is only the implicature ‘I cannot take you home’. Or suppose you tell me that “He’s arrived”; I may be at a loss to understand who is meant by “he”, but still understand that some male person has arrived. In short, when the analysis of a turn contains M(S) and N(H), as in [1.1] and [1.2], or M1(S) and M2(H), as in [1.4], the hearer will usually have recovered correctly some aspects of the message and will have failed to recover others. Instances of ‘non-understanding’, or complete failure to recover anything, will be those where the hearer fails to hear the utterance correctly or, due, perhaps, to lack of knowledge of the language, fails to understand the literal meaning of the speaker’s utterance. We conclude that there is no need in our system for a theoretical concept ‘non-understanding’. Instead, a full analysis ought to specify which aspect of the utterance has been misencoded by the speaker or has been incorrectly retrieved by the hearer.

580 Izchak M. Schlesinger and Sharon Hurvitz

In other words, in addition to the structural analysis of misunderstandings discussed in the preceding sections, we have to specify for each discrepancy involving M and N — e.g., M(S) and N(A) or M(S) and N(H), M(A) and N(H) — or M1 and M2, where the misunderstanding is located. In the next section we propose a classification based on the phase in the process of understanding that has been disrupted or the aspect of the production process that has been affected. Previous attempts in this direction are those by Falkner (1997) and Dascal (2003: 303–321), but the following proposal departs from theirs in some respects. 9. Misunderstandings — WHAT went wrong? The English language distinguishes between failures in various processes that may result in misunderstanding: e.g., between mispronouncing, mishearing, and misinterpreting (see Dascal 2003: 293 for a fuller list of terms with ‘mis-’). Grimshaw (1980) distinguishes between perceptual, linguistic, and pragmatic processes. The various processes involved in understanding a message correspond to what Dascal (2003: 304ff) calls ‘layers of significance’. Dascal views these layers as embedded one within each other, with speech perception forming the innermost layer. In the following, a list of layers is proposed, at each of which a speaker-caused misunderstanding or else a hearer-caused misunderstanding may occur.9 The message, M, that our formulas refer to — in M(S), M(A), and M(H) — includes the totality of these layers. 1. Speech perception. The hearer must perceive the speech sounds produced by the speaker. Due to environmental noise, a faulty line in a telephone conversation, or because he is hard-of hearing, etc., the hearer may mis-hear or fail to hear the utterance. Similarly, a speaker-induced misunderstanding may result from the speaker mispronouncing a word (stuttering, committing a Spoonerism, etc.). The same applies of course, mutatis mutandis, to written communication (to the writer and reader, respectively). 2. Decoding of the literal meaning. The hearer makes sense of the perceived speech sounds by applying the semantic and syntactic rules of the relevant language. When he deploys the wrong rules — whether due to insufficient knowledge of the language or due to other factors, like lack of attention — a misunderstanding may ensue. A speaker-induced misunderstanding may result from the speaker’s failure to adhere to what Grimshaw (1980) refers to as ‘linguistic rules’. 3. Interpreting. Understanding the meaning of words and analyzing the syntactic structure are not sufficient for understanding. In the following, we list some kinds of interpreting that are required if communication is to be considered successful:



The structure of misunderstandings 581



a. The meaning of ambiguous or vague words and the meaning of syntactically ambiguous sentences have to be recovered. In Grimshaw’s (1980) terms, this, too, is a linguistic process; the following are pragmatic ones: b. The hearer has to determine the illocutionary force of the utterance. c. The hearer has to understand the ‘point’ of the utterance (Dascal 2003: 308). d. The hearer has to recover what Falkner (1997: 140–150) calls the intended ‘Modifikatoren’ of the speaker’s utterance. The speaker may have been ironic, she may have spoken jokingly, metaphorically, or have been parodying or quoting someone; she may have indulged in overstatement or have been polite. The hearer may then misunderstand her and take her to have been serious, literal, and so on. Or else, the speaker may have been serious, literal, and so on, and the hearer may wrongly attribute a different ‘Modifikator’ to her utterance.  To retrieve this layer, the hearer has to consult the context of the utterance, the ‘common ground’ of the interlocutors, and general world knowledge, and in attempting to do so he may make mistakes. Speakercaused misunderstandings may also originate at this level. For instance, the speaker may know only one of the meanings of an ambiguous word, or one of the ways a construction is used, while in the particular context of her utterance these will normally be construed differently. A type of a speaker-caused misunderstanding at the interpreting phase is exemplified by (2) (in Section 2). 4. Retrieving implicatures. This layer concerns implicatures proper — as distinct from the ‘explicatures’ or ‘implicitures’, as they are called by some writers — that belong to the previous layer. The hearer may fail to retrieve the implicated proposition as illustrated in (1), or retrieve the wrong one (or pretend to retrieve the wrong one), as in (4) and (8). A speaker-caused misunderstanding in respect to implicatures occurs when the speaker formulates her utterance in a way that prevents the intended implicature from getting across. In (9), for instance she does so purposely. Let us illustrate layers 1–4 with a short dialogue: (11) Cyril: Can David have your car tonight? Bernie: Tha-that’s a problem. The t-t-t-tire has a flat. The other tire too.

Observe, first, that on hearing Bernie’s utterance, Cyril’s comprehension process may go wrong at any one of the four layers mentioned in the preceding: Speech perception: Due to Bernie’s faulty enunciation, Cyril may fail to make out which words have been uttered.

582 Izchak M. Schlesinger and Sharon Hurvitz

Decoding the literal meaning: Being British, Cyril may wonder what a “flat” has to do with cars (the British-English word appropriate here being “puncture”). Interpreting: It doesn’t take much to realize that the tires Bernie refers to are the tires of his car. However, there is a, albeit very slight, possibility that Cyril does not catch on here (perhaps he is inattentive). Retrieving implicatures: The implicature of Bernie’s utterance is of course: David cannot have the car. To retrieve this, Cyril has to draw on his general knowledge: one tire can be replaced by the spare tire, but when two tires are punctured, you have to take the car to a garage. Cyril’s ignorance in this respect may result in his failing to draw the implicature. The types of misunderstanding dealt with so far are those usually discussed in the literature. But there are other things, too, that the hearer may do with an utterance. If we regard any disruption of the communication process as a misunderstanding, we may identify additional types (rarely dealt with, perhaps because less damaging to the interaction). Let us look now at these additional types of disruption. 5. Appreciating the form of the message. Full comprehension involves not only identifying the informational content of the message, but also appreciating the way the message has been phrased: in what language the message is formulated and in what ‘register’, whether the speaker’s diction is clear, or — as in our example (11) — whether he stutters. Suppose, for instance, the speaker indulges in rhyming, or makes a pun, and the hearer fails to notice this. Or the speaker makes her point by using a quote, and the hearer does not appreciate this fact. Then something gets lost in the interchange even though the message content may have been fully recovered by the hearer. Conversely, when the speaker inadvertently makes a rhyme, the hearer may notice what the speaker did not intend. These, then, would hardly be instances of a perfectly smooth interaction. The pre-theoretical notion of misunderstanding perhaps does not include such cases, but for our purposes it is appropriate to stretch the term so as to include them. 6. Resonating to the message. Frequently, an utterance releases in the hearer some kind of attitudinal response: he may agree or disagree with what has been said, be outraged or pacified by it, view the speaker as having been sincere; fraudulent, impolite, obsequious, and so on. In (11), Cyril may understand the implicature and react to it emotionally (thinking: ‘The bastard! He drives me mad.’). When Bernie’s motives were such that they did not merit such resonating, we have a kind of misunderstanding. 7. Monitoring the discourse. In conversation, utterances often occur whose function is not to impart information, but to ‘steer’ the discourse: for instance,

The structure of misunderstandings 583



indications of an intention to intrude (“But…”), or so-called ‘continuers’ (like “uhuh”). When the hearer fails to ‘understand’ these signals, a misunderstanding, of sorts, has occurred. Further, in regard to a regular, informative utterance, the hearer has to understand its discourse function: is it a reply to a question — as in (11) — or has the speaker embarked on a new subject, etc. 8. Inferring: Hearers make inferences from messages, and these may diverge from those intended by the speaker. Sherlock Holmes may tell Watson a certain fact, intending him to draw from it a certain conclusion; if Watson then arrives at a different conclusion, there has been a disruption of communication; Watson has misinferred from the detective’s utterance. In (11), Cyril may make the inference ‘He doesn’t want to let David have the car’, and may be dead-wrong: Bernie may have loved to lend the car (and perhaps his stutter partly reflects his intention to have Cyril think that he is embarrassed by his failure to comply). Note that we are talking here only about intended inferences — what the speaker’s utterance may indicate (using Dascal’s [2003: 39] terminology). Here is another example: (12)

A1: Hullo, I am calling about your reservation for the flight next Thursday. B: Oh, did they cancel it? I really can’t afford to put it off much longer. A2: No, I am calling to remind you that your ticket hasn’t been paid for yet. (after Falkner 1997)

The opening turn, “I’m calling…” cannot be said to involve an implicature concerning what the call is about; but B mistakenly infers from the fact that the company calls her that she is going to be told about a cancellation. 10. Summary Our analyses have shown that misunderstanding is a multifaceted concept. The conceptual framework proposed in this article comprises several facets of interpretation: the message as understood by the speaker, the hearer, and the analyst, the speaker’s intention as to how the hearer is to understand the utterance, the hearer’s perception of the speaker’s intentions, and the way an interlocutor pretends the message to have been understood by herself or her addressee. These constructs afford a classification system for misunderstandings. A proposed orthogonal classification is based on the layer of significance that fails to be communicated correctly. The analyses presented here involve viewing ‘misunderstanding’ as a theoretical concept that diverges somewhat from that denoted by the pre-theoretical term ‘misunderstanding’, as it is generally understood.

584 Izchak M. Schlesinger and Sharon Hurvitz

Notes *  Thanks are due to Moshe Anisfeld, Marcelo Dascal, Anita Mittwoch, and three anonymous reviewers for very helpful comments. 1.  These examples illustrate that a misunderstanding may become apparent only in a subsequent turn. 2.  Cf. the notion of a neutral observer in controversies envisaged by Leibniz (Dascal 2003: 281). 3.  The three concepts M(S), M(H), and M(A) have a parallel in three approaches to the interpretation of literary texts. One school holds that the meaning of a text lies in what the author intended it to be, whereas according to a different, more modern view, it is the reader, rather than the author, who creates meaning. But scholars of the New Criticism school hold (to put it in a simplified form) that the meaning resides in the text, and not in the author’s intention or the way the reader reads it. 4.  Some writers, though, would reserve the term ‘misunderstanding’ for the latter. Dascal (2003: 294) states that they are a “communicative phenomenon typically belonging to reception” (emphasis added). As will be seen in the following, one and the same conceptual apparatus permits the classification of speaker-caused as well as hearer-caused misunderstandings, and we propose therefore that the theoretical concept of misunderstanding should include both. 5.  As are misunderstandings where the two errors cancel each other out, so that N2(H) in the above formula is replaced by M(H). 6.  This and other examples which are quoted in the present paper from Falkner (1997) are reallife examples. 7.  Normally, M1 and M2 will not be equally salient: There is a gradient from a totally unacceptable interpretation (taking context into account) and a totally acceptable one. 8.  This, rather than M(S), is the concept of intention that figures in Grice’s (1969) definition of non-natural meaning. Observe also that our M(A) has no parallel in Grice’s system: what Grice (1968) calls “applied timeless meaning of an utterance type” does not include the implicatures which an ‘analyst’ retrieves from the utterance. Further, since he is not concerned with misunderstandings, Grice has no need for M(H). 9.  A caveat is in order here. No claim is being made concerning the sequence of the procedures in the comprehension process. This is a psychological issue that does not concern us here.

References Bazanella, C. and Damiano, R. 1999. “The interactional handling of misunderstandings in everyday conversations”. Journal of Pragmatics 31: 817–836. Blum-Kulka, S. and Weizman, E. 1988. “The inevitability of misunderstandings: Discourse ambiguities”. Text 8: 219–241.



The structure of misunderstandings 585

Dascal, M. 2003. Interpretation and Understanding. Amsterdam: John Benjamins. Falkner, W. 1997. Verstehen, Mißverstehen und Mißverständnisse. Tübingen: Niemeyer. Grice, H.P. 1968. “Utterer’s meaning, sentence meaning and word meaning”. Foundations of Language 4: 225–242. Grice, H.P. 1969. “Utterer’s meaning and intentions”. Philosophical Review 78: 147–177. Hockett, C.F. 1986. “Grammar for the hearer”. In G. McGregor (ed), 49–68. Humphreys-Jones, C. 1986. “Make, make do and mend: The role of the hearer in misunderstandings”. In McGregor (ed), 105–126. Humphreys-Jones, C. 1987. “The structure of misunderstanding”. In R. G. Reilly (ed), Communication Failure in Dialogue and Discourse. Amsterdam: North Holland, 25–33. Locke, J. 1690. An Essay Concerning Human Understanding. (Ed. P. H. Nidditch, 1975). Oxford: Clarendon Press. McGregor, G. (ed). 1986. Language for Hearers. Oxford: Pergamon Press. Nerlich, B. and Clark, D. H. 2001. “Ambiguities we live by: Towards a pragmatics of polysemy”. Journal of Pragmatics 33: 1–20. Schegloff, E.A. 1987. “Some sources of misunderstanding in talk-in-interaction”. Linguistics 25: 201–218. Schegloff, E.A. 1992. “Repair after next turn: the last structurally provided defense of intersubjectivity in conversation”. American Journal of Sociology 97: 1295–1345. Taylor, T. 1986. “Do you understand? Criteria of understanding in verbal interaction”. In McGregor (ed) Language for Hearers, 81–103. Trognon, A. and Saint-Dizier, V. 1999. “L’organisation conversationelle des malentendus: Le cas d’un dialogue tutoriel”. Journal of Pragmatics 31: 787–815. Weigand, E. 1999. “Misunderstandings: The standard case”. Journal of Pragmatics 31: 763–785.

Authors’ addresses Izchak M. Schlesinger Department of Psychology Hebrew University of Jerusalem Jerusalem 91905 Israel

Sharon Hurvitz Department of English Hebrew University of Jerusalem Jerusalem 91905 Israel

[email protected]

[email protected]

About the authors Izchak M. Schlesinger is Professor Emeritus of Psychology at The Hebrew University of Jerusalem and first incumbent of the Joseph H. and Belle Braun Chair in Psychology. He is the author of several books — among them: Cognitive Space and Linguistic Case (1995) and The Structure of Arguments (with T. Keren Portnoy and T. Parush, 2001). Sharon Hurvitz received her Master’s degree in English linguistics at the English Department of the Hebrew University of Jerusalem in 2005, and was awarded the Dean’s grant in 2006. She is now working on her PhD at the same department, where she has also been teaching for the past few years.