Resisting the legitimacy of the question: Self-evident answers to questions about sources of knowledge in police interviews with child witnesses

2.1 Knowledge in everyday question sequences

If people ask questions for information in everyday life, they generally propose that the addressee of the question knows more about that particular topic than the person who asks the question (e. g. Heritage 1984a: 250; Heritage and Clayman 2010: 25; Heritage 2012a). That is, a question suggests that the addressee of the question is better able to provide the requested piece of information (K+) than the person who is asking the question (K-) (Heritage and Raymond 2012). “Suggests” is a keyword here. It distinguishes between the language that proposes a difference in knowledge, and the longer lasting “actual” knowledge that speakers have. The epistemic situation proposed in the language is often referred to as “epistemic stance”; the more stable “real” knowledge is referred to as “epistemic status” (Heritage 2012a, 2012b). Typically, if someone asks a genuine question, the epistemic stance is congruent with the epistemic status. That is: the person who asks the question cannot provide the requested piece of information (Heritage 1984a: 250; Heritage 2012a: 377; Heritage 2012b; Heritage and Raymond 2012: 3). It is important to differentiate between stance and status because the notion of epistemic stance makes room for negotiation. Proposals about what someone knows and does not know can be misunderstood, challenged, disputed and confirmed (e. g. Pomerantz 1984; Heritage and Raymond 2005; Fox 2001: 171).

Negotiations about how much and which knowledge a question presupposes take place throughout the question sequence, i. e., by posing the question (first position), by giving an answer (second position), by taking up that answer (third position) and possibly beyond that. The following excerpt from everyday interaction, also used by Heritage (2012b: 19), illustrates that point nicely:

In line 11, Mom asks whether Russ knows who is going to the meeting. This turn has an interrogative grammar, but it can be interpreted as doing a pre-telling (Terasaki 2004) if one figures that Mom already knows who is going to that meeting. This is in fact how Russ treats the question (line 12): he inquires who is coming and thus gives a “go ahead” for his mother to tell who is going to the meeting. His answer thus displays that he interprets his mother’s utterance as coming from a knowing epistemic status (K+), and therefore doing something different than questioning. By requesting his mother to tell who is coming to the meeting, he also takes a stance of being unknowing himself (K-). In third position (line 13) it appears that Ross has misunderstood his mother’s epistemic status. In line 13, Mom claims having no prior knowledge (K-), which retrospectively clarifies that the utterance in line 11 should have been understood as a question, rather than a pre-telling (Heritage 2012b: 19–20). This clarification corrects Ross’ presumptions as they are proposed in line 12. Ross then starts answering the initial question as a request for information (line 15–16), which proposes a relatively knowing stance (K+). This is remarkable when compared with the initial unknowing stance he has taken in line 12. The excerpt shows that what we think the other person knows, is critical in understanding if an utterance is a question (i. e. requests information) or a telling (i. e. conveys information) (Heritage 2012b). Moreover, it shows that what is known is subject to interactional negotiation and therefore implicit and explicit knowledge claims might change throughout sequences.

Extract 1 further more illustrates that questions suggest a knowledge asymmetry. This is not only a description of what happens when a speaker asks a question, but it also has a normative aspect: speakers are supposed to orient themselves towards this knowledge asymmetry. This is observable in interaction when speakers “ask for the obvious” and they or their interlocutors treat that as an accountable action. For example, Stokoe and Edwards find that police officers frame some questions as “silly”, like “Um (1.4) may sound a bit silly bu- uh do y’know whose window it is” (simplified from Stokoe and Edwards 2008: 92). The account “may sound a bit silly” displays that the question transgresses an interactional rule “don’t ask things you already know” (Stokoe and Edwards 2008) and therefore provides evidence that this rule exists. Similarly, answers can hold questioners responsible for questions that ask for known information. “Of course” and “absolutely” in responses treat a question as unaskable based on for example general knowledge or previous discourse (Stivers 2011); the change of state token “oh” in response-initial position treats the question as unexpected or inappropriate and suggests that the questioner should have known better (Heritage 1998; see also Sidnell 2012). The moral obligation not to ask for known information mirrors the interactional rule about telling “don’t tell others what you figure they already know” (Jefferson 1992: xv; Sacks 1992: 438). Speakers orient to this rule by practices such as pre-tellings: “Didju hear the terrible news” (example from Terasaki 2004: 184). A negative response ensures that the speaker will not be giving redundant information (Terasaki 2004: 183–193). Addressees of a telling can also hold the teller morally accountable for transgressing an interactional rule by interactional sanctions. For instance, re-tellings to the same person, could be interactionally sanctioned by the response “you already told that” (Sacks 1992: 438).

2.2 Sources of knowledge in everyday interaction

Sources of knowledge play an important role in how speakers’ position themselves epistemically (Sidnell 2012). In many languages sources of knowledge, or evidentiality, must be provided in order to produce a grammatically correct utterance (e. g. Chafe and Nichols 1986; Aikhenvald 2004). In other languages, like English and Dutch, providing such sources is optional (e. g. Fox 2001: 168; Sidnell 2012: 128). Pomerantz writes: “Recipients may infer the access or bases speakers have for making assertions. (…) And routinely if the source is what one would normally infer, it is not explicitly described” (Pomerantz 1984: 610–611). Consequently, when speakers add a source of knowledge to their assertions, they do something specific (Pomerantz 1984). Consider for example:

In example 1a “according to this article’ provides a source of knowledge and simultaneously distances the speaker from the assertion that “there are over a hundred thousand”. The speaker thus only takes responsibility for correctly quoting the article (Pomerantz 1984). In example 1b “sounds like” presents the evaluation as based on what the interlocutor has reported about a particular event, rather than on direct access to the event. The reserved assessment of the event thus indexes that the other person is more entitled to assessments, in this case because of his or her direct experience and knowledge of the event (compare e. g. Raymond and Heritage 2006; see also Stivers et al. 2011). Using evidential or epistemic marking is, therefore, not just a matter of indexing that something is “not certainly, definitely, and unproblematically extablished” (Pomerantz 1984: 608), downgrading the certainty or representing an actual cognitive state. It also orients to the relationship between interlocutors (Fox 2001), the sequential positioning (Fox 2001) and the knowledge of one interlocutor relative to the other (Sidnell 2012; see also Kim 2005). Another point to take from the examples relates to the sequence in which they may occur. The first example could be taken from a discussion about refugees and the second one from reporting about a holiday. In such sequences, the main activity is not establishing “who knows what and how”. Instead, the speakers use sources of knowledge to manage their mutual epistemic relationship “en passant” (Sidnell 2012).

Similarly, asking for a source of knowledge does more than merely ask for information about a cognitive state. Asking for evidence, including sources of knowledge, implies that the validity of a given assertion is not taken for granted and that there is some degree of doubt about what was asserted (Pomerantz 1984). Moreover, whereas much negotiation about who knows what and how happens en passant, asking for a source of knowledge topicalizes the source of knowledge and it highlights the accountability of the other speaker for the correctness of his or her assertions.

From these considerations it follows that the questions about sources of knowledge that police officers are instructed to ask are marked. The interview guidebook does not prescribe that children should be informed about the background of these questions (Dekens and Van der Sleen 2013) and in our interview data police officers never explain to children why they need to ask these questions about sources of knowledge. A relevant question, then, is how children deal with questions about sources of knowledge.

5.1 Responses presented as self-evident

Excerpt 2 (below) shows one way of presenting an answer as self-evident: using “nou” (“well”) and “gewoon” (“just”). The excerpt is taken from an interview with a 6-year-old boy who was playing in the attic with his teenage cousin. The reason for the interview is that the boys allegedly have rubbed body parts to each other: their noses, bellies, buttocks and willies. Just before the extract, the police officer asked whose idea it was to do this. Throughout this paper “P” indicates the police officer’s speech and “K” indicates the child’s speech.

The question about the child’s source of knowledge is produced in lines 3 and 5–9. The question has several versions that move from making the cousin responsible for the activity (line 3: arguably heading for “that he wanted to do that”) and emphasis on what the child (the singular form of “you”) “had to do” (line 5), to presenting the activity as a joint activity by using a plural form “jullie” (“you”, line 7). The child does not respond directly (line 8) and starts frowning during the production of the question (line 5), displaying difficulty to understand the question or possibly “thinking”. The police officer further clarifies “that” (line 7) with an incremental “bottom to bottom” (line 9).

The child first delays his answer with 0.7 seconds (line 10) while still frowning, which suggests an ongoing struggle to respond and to make sense of the question. The delay is followed by “nou” (“well”) which constructs the upcoming answer as not straightforward (Schegloff and Lerner 2009), as a shift in perspective or departure from constraints of the question (Heritage 2015). The modified repeat of the question (line 12) also characterizes the question as somehow problematic (Bolden 2009). In the next line, the child proceeds the prefacing of the answer with “nou gewoon” (“well we just did”; line 13). “Well” has also been noted to contest the relevance of a question (Heritage 2015: 93) and the Dutch “nou” serves this job too in this fragment. The use of “nou” also seems to be colored and reinforced by “gewoon” (“just”). “Gewoon” presents the answer-to-come as common sense and therefore as self-evident, that is, not needing other evidence (Fox 2001: 173). The answer preface in line 12 thus proposes that the answer was readily available for the police officer to infer and resists the question as a legitimate request for information (for different, yet also resisting, uses of “gewoon” see Lamerichs et al. 2015; Van Charldorp 2011). Despite this resistance, the child attempts to provide an answer.

Another resource for presenting an answer as self-evident is intonation. This is shown in extract 3, taken from an interview with a 9-year-old girl. Her parents are divorced and she is living with her mother and her mother’s new partner; the girl regularly visited her father during weekends. She claims that her father has repeatedly abused her during these weekends. The girl has drawn a map to point out where she and her father were in the bedroom.

The girl confirms the police officer’s summary of the physical situation in line 3 with “<hm hm:>”. She also orients to more talk by the police officer by producing only a minimal confirmation, assuming a listener’s role (Gardner 2001). The stretching of “<hm hm:>” makes it sound slightly reluctant. The police officer indeed produces further talk in the form of a question about how the girl knows what she had to do (lines 4–5).

The girl substantially delays her response indicating some sort of trouble whilst she keeps looking at the drawing but without visible signs of being upset with the question (line 6). She maintains this pose until the end of line 15. In line 7, the girl responds “•pt # ↑because he has already explained that°,”. The emphasis on “omdat” (“because”) combined with the noticeable higher pitch as well as the emphasis on “uit” (“ex” from “explained”) present the answer as self-evident. The intonation of the response can even be heard as impatient but the lowering volume by the end of the utterance softens this. The intonation thus presents the answer as something that was obvious. This implies that the answer is something the police officer should have understood without asking the question.

It might be argued that the child is anticipating the follow up questions and that she is dreading to say more about the traumatizing things she had to do. While this could be the case, the “•pt # ↑because he has already explained that,” is still hearably produced as obvious. The continuation of the sequence also points in the direction of an anticipated repeated activity causing trouble, rather than the sensitivity of the topic. This becomes especially clear after line 15. In line 16, the girl starts smiling and then looks at the police officer with laugh on her face (line 17; the laugh is not hearable). The police officer treats the beginning of this laughter as displaying that the child has already told what “sucking off” means by accounting for her question as a repeated question. The hand gesture (line 17) physically “holds of” the complaint about the repetition. The child confirms the asserted repetitive nature of the question by making clear that they have dealt with this issue before and even reinforces the problem by correcting that is already the third time the police enquires about this topic (rather than the second; line 18), thus marking a transgression of the interactional rule “don’t ask things you already know” (Heritage 1998; Sidnell 2012; Stivers 2011; Stokoe and Edwards 2008). The police officer aligns with the child’s complaint by admitting that the repetition of the question is “a bit silly” (line 20) and by confirming that the child is allowed to say this (line 21). This response acknowledges that the repetition of the request for a definition is an accountable action. Although the topic probably does not make the child more eager to talk about it once more, in what happens after the self-evident answer she treats the repetitiveness of some questions as problematic, rather than the topic of these questions.

“Nou”, “gewoon” and intonation are common ways of presenting answers as self-evident in response to questions about sources of knowledge found in the dataset, and they often co-occur. The next example shows another way of presenting the answer as self-evident that is much less common in our collection of questions about sources of knowledge. The extract emerges at the very end of the interview with a 6 or 7-year-old boy ^[14] about having been licked on his teeth by an unknown teenage boy in the park. Possibly the licking refers to an attempted “French kiss”. The boy has reported during the interview that he has talked about the event with a girl named “Dine”. The police officer asks for further details about this girl near the end of the interview while they are playing a game, presumably to collect information about a potential witness or to get on tape who this girl is.

The police officer reintroduces the girl’s age that has been established in the 13 lines that have been omitted here (line 19), eliciting the confirmation in line 21 by using rising intonation. Then she asks how the child knows the girl’s age (line 22). This question is also met with a substantial silence (line 23) indicating trouble, similar to extract 2 (line 10) and extract 3 (line 6). The child stares in the police officer’s direction during this pause. Next, the “nou” (“well”) preface is combined with increased volume in “KEn” which presents the answer as obvious and possibly also impatient. The child witness thus presents the fact that he knows this girl as readily available to the police officer. A feature that distinguishes this response from extract 2 and 3 is that the content of this response also works as a resource to present the answer as self-evident character. The claim that the boy knows the girl serves as an account for how he knows her age. This response presents someone’s age as part of the package of things that you know when you know someone and employs that “package” to resist the legitimacy of the police officer’s question. “Knowing someone” thus works as a good enough ground to know someone’s age. Therefore, the child suggests that the police officer already should have understood how he knows the girl’s age.

The self-evident answers discussed above contest the necessity and legitimacy of asking the question about the source of knowledge. The different responses exploit different ways of how the police officer could know. The preface in excerpt 2 frames the answer as obvious in general, excerpt 3 suggests that the police officer should have known because of what was said earlier in the encounter, excerpt 4 uses knowledge related to social relations. Excerpt 2 and 4 are in line with Pomerantz’ observation that sources of knowledge can normally be inferred and are usually left implicit (1984: 610–611). The self-evident response in excerpt 3 shows that issues that already have been addressed should not be repeated. Answers presented as self-evident thus hold the police officer accountable for transgressing a combination of two “rules”: (a) you already should know how I know, based on the sequence or everyday inference rules concerning sources of knowledge, and (b) if you already know or can know my source of knowledge, don’t ask (as a variation of Sacks’s “don’t tell people what you figure they already know” 1992, vol 2: 438; compare Stokoe and Edwards 2008: 92). The self-evident answers thus not only indicate that police officers could know, but also that they should know (compare Stivers et al. 2011). This makes sense if we recall that sources of knowledge usually can be left implicit (Pomerantz 1984).

It is notable that the extracts show that the children attempt to provide an answer to the question about their knowledge. This is the usual pattern within our data: even if children present their answers to questions about sources of knowledge as self-evident, they attempt to provide an answer anyway. It is interesting to contrast these responses with how children could respond. They could have spent a whole turn on contesting the question and refusing to provide an answer, e. g. by saying something like “why do you need to ask this” or “aren’t you listening?”. By attempting to answer instead, children visibly orient to the moral obligation to provide an answer about the source of knowledge that was created by the police officer’s question. The need to provide an answer possibly becomes even stronger given the asymmetric power distribution between the adult law enforcement representative and the child. The dispute about how legitimate the question is, takes place in a relatively implicit way: using discourse markers, by using intonation and exploiting knowledge about for example social relationships. Answers presented as self-evident therefore show that the children are dealing with a dilemma. On one hand, the police officer first pair part (the question) has created a moral obligation to respond. On the other hand, the children’s mundane inference system indicates that this question was unjustified. Doing two things at the same time – answering and disputing – enables children to simultaneously deal with both the question and, from their perspective, its unnecessary and unjustified character. Although children skillfully deal with both issues, the fact that they treat the question as causing a dilemma makes apparent that they struggle to make sense of the question and that the institutional purpose of this question is not always treated as immediately evident to them.

5.2 Police officers’ responses to self-evident answers

Interestingly, police officers usually ignore the resistance offered by self-evident answers. They tend to acknowledge the answer with an answer repetition, a sequence closing “okay” (compare Gardner 2001: 54; Beach 1995; Gaines 2011), a follow-up question or a combination of two or more of these options. Extract 3 provides an example in lines 9 and 10 (for convenience repeated as extract 3a):

In the extract, the police officer continues the line of questioning without attending to the intonation that produces the child’s answer as something that the police officer should have thought of herself. Sometimes, however, police officers’ uptakes do a bit extra, for example in extract 5. The extract is taken from the same interview as extract 4, about the alleged French kiss.

The police officer asks how the child knows the man’s age. The boy presents his answer as self-evident using “nou” (“well”) and emphasis on “said”. The police officer receives the answer with the “news receipt” “oh” (Heritage 1984b). This receipt marks a “change of epistemic state” which retrospectively shows that the child’s answer was informative, even though the child presents the answer as obvious. The police officer maintains not having known the answer and justifies the question as not having such an obvious answer. The police officer thus pushes back the self-evident presentation of the child’s answer.

Police officers can also use other resources to push back on the self-evident answers. The next excerpt shows how “okay” may be transformed into a news receipt. The excerpt is taken from an interview with a 6-year-old boy. He has been approached by an unknown man who showed his private parts. The police officer tries to find out how long the man’s coat was, establishing if the child could indeed see the man’s private parts. The child compares the length of the coat to his mother’s coat, that she was also wearing when bringing her son to the police station.

The question about how the child could see the man’s willy (line 4) is framed as a puzzle: if the man’s coat was that long, how could the boy have seen the man’s willy? In line 6, the child responds to this challenge by presenting his answer as self-evident, using “nou”, “gewoon” and strong intonation in “broek” (“pants”). The police officer responds to that with “okay”. Yet, the police officer uses a strongly rising intonation to do something more than just acknowledging. In this case, the intonation adds an element of a “news receipt” (compare Heritage 1984b: 307–312). Consequently, “okay” treats the answer as informative and the question as legitimate, despites the child’s presentation of the answer as self-evident. ^[15]

The point of justifying the question about the source of knowledge becomes clearer if one considers what the police officer could have done. One option would have been to acknowledge in third position that the question was silly. For example, the police officer could have said in extracts 4 and 5: “Yes I know it sounds silly, but I just need to be sure.” The police officer would then have aligned with the child’s implicit proposal that the answer was obvious and that the question was indeed performing an accountable action (compare extract 3 lines 16–23 in response to the child’s laughter after the question to “again” explain what “sucking off means”; compare Stokoe and Edwards 2008: 92). Such an uptake would have established that the question is asked for other reasons than collecting information, for instance producing this information explicitly for the record. Instead, the uptakes “oh” and “↑↑ok↑↑ee?” following self-evident answers to the questions about sources of knowledge retrospectively testify that the police officer did not know, and therefore accounts for the question as being genuine and justified.

A more explicit and rare way of legitimizing the question can be found in the following excerpt:

In line 7, the child responds to the question about the source of knowledge (lines 5–6) by smiling and answering “I was sitting th↑ere, (hh)”. The high pitch in “there” combined with the laughter presents the answer as self-evident. This implies both that the police officer could and should have already understood that the girl was there, and that this should have been enough ground to understand how the child knows. The police officer repeats the answer and acknowledges it (lines 8–9). In overlap with the police officer’s “okay”, the child produces another source of knowledge that is available to the police (the child elsewhere says that her father has given a printout of the chat to the police). It is also marked with “ook al” (“also” or “already”) which presents this source as “yet another source”, highlighting that the police officer has even more ways of understanding. In line 12 the police officer proceeds her uptake of the child’s answer with “that wasn’t entirely clear to me” (line 12). This makes explicit that she did not understand before, and that it is clear to her now. This is also an account for asking the question: it was a missing piece of information earlier and the question was, therefore, a genuine information seeking question.

In sum, police officers generally ignore the self-evident presentation of children’s answers that question how justified questions about sources of knowledge are. In cases where they do orient to that negotiation, they rather retrospectively account for the question as a real question about something they did not know. By doing so, they implicitly align with the second rule children orient to in their self-evident answers: (b) if you have already know my source of knowledge, don’t ask. However, they resist the suggestion that they could have inferred or known how the child knows.

5.3 An institutional inference system

The question is why police officers apparently do not infer or know what can be evident from the child’s perspective. A possible explanation is that police officers inferences are limited by institutional constraints that dictate that they should not conclude too quickly that they understand how the child knows. This explanation is supported by the following excerpt taken from an interview with an 8-year-old boy who has been exposed to “strange” and “ugly” films by his grandfather, possibly referring to pornographic material based on further description in the interview (data not shown). The boy claims that his grandfather had received those films by email. In the excerpt, the boy has gone to the bathroom and the police officer who monitors the interview (P2) uses the opportunity to give feedback to the interviewing police officer.

In lines 1–2, P2 suggests that the interviewing police officer should ask the boy how he knows about the emails that his grandfather receives. P2 seems to be introducing another question in line 8, but then restarts by accounting for why P should ask for a source of knowledge. She proposes that “we” fill in that the child knows about the emails because that has been talked about (line 9), but then suggests that there could be other sources of knowledge (line 10). P2’s account treats “how the child knows” as something that cannot be just inferred or “filled in”, and as something that is not known until the child has explicitly stated what the source of knowledge is.

The interviewing police officer orients to that same rule. First he acknowledges the advice (line 12), then he denies that he fills in how the child knows (line 13) and announces the course of action that he will take after the feedback session (line 14) i. e. asking how the boy knows about the email. Notably, the police officer uses “gewoon” (“just”) (line 14), and presents asking the source of knowledge as the normal way of doing things rather than filling in how the child witness knows. Both the denial of “filling in” sources of knowledge in line 13 and the “ordinariness” of asking for a source of knowledge display that this is the normal way of doing it and that the police officer is aware of that. Asking for sources of knowledge is therefore not only oriented to as something the police officers have to do in order to facilitate assessments of reliability, it comes forward as a normative rule that prescribes that police officers should be careful not to infer a source of knowledge too quickly, that is: before it is made explicit. At the same time, the fact that the police officer is urged to ask a question about sources of knowledge suggests that this inference system may not always come naturally and needs monitoring.