The Interactional Costs of “Neutrality” in Police Interviews with Child Witnesses

ABSTRACT

This paper concerns the interactional dilemma between displaying affiliation and doing being neutral. This dilemma is highly salient in police interviews with child witnesses where interviewing guidelines encourage police officers to take a neutral stance to avoid steering children’s stories. In this article, we use conversation analysis to analyze childrens’ volunteered accounts of their own role during the alleged offense, e.g., how they resisted. Such accounts make relevant affiliative uptakes such as approval, disagreement, or reassurance that may be seen as nonneutral. Hence, these accounts raise interactional dilemmas for police officers: Should they do what is interactionally relevant or follow the guidelines? Our analysis shows how police officers display and deal with this dilemma and that children may add to it by pursuing something more than neutralistic uptakes. The upshot of this analysis is that attempting to be neutral in interaction may cause apparently undesirable interactional difficulties. The data are from the Netherlands.

Investigative interviews with children are a highly regulated type of interaction. In the Netherlands, where the data presented in this article were collected, guidelines for interviewing children can be found in the Manual the Child as a Witness (Dekens & Van der Sleen, 2013). The aim of such guidelines is twofold: to enhance the reliability of the testimony and to support the child as much as possible. One salient guideline that targets the former aim is that police officers are encouraged to conduct the interview as neutrally as possible:

Because fact finding is key, the interview must be conducted in the most neutral way possible. (…) In this context, neutral means: without influencing the witness’s testimony, neither in posture, mimic and intonation, nor in the way questions are formulated. No matter the means, influencing must be avoided. (Dekens & Van der Sleen, 2013, pp. 13–14, our translation).

An instructor at the police academy explained that police officers should not assume that something has happened because the child’s story is still under investigation, and it is the police officer’s task to find the truth.¹ The phrase “the most neutral way possible” in the quote acknowledges that being fully neutral is not possible. Elsewhere in the Manual this is made even more explicit by stating that “in practice, it is difficult or maybe even impossible to do an interview in an entirely neutral way” (Dekens & Van der Sleen, 2013, p. 47). Nevertheless, being neutral is prescribed as the police officer’s aim and is mentioned as important by police officers and trainers (Rassin & Van Koppen, 2002, p. 26).

In this article we use conversation analysis (CA) to examine moments in police interviews with child witnesses in which the police officer’s neutrality is at stake. We focus on children’s volunteered accounts of their own role during the alleged offence (cf. Robinson, 2016), and on how police officers respond to these accounts displaying a dilemma between “being neutral” versus affiliating with the child’s account. Finally, we show how children sometimes treat uptakes that “do being neutral” as insufficient by pursuing something more. This provides interactional evidence for the police officers’ dilemma. Before we present our analysis, we discuss some previous literature to provide a context for our analysis.²

Accounts in legal settings

Participants in legal settings have been found to orient to accountability in the sense of liability. This is perhaps unsurprising, as Edwards and Potter point out, because the whole point of courts is to decide if responsibility can be attributed or not (Edwards & Potter, 1992, p. 159). Therefore, many suspects and defendants present their role in a way that denies responsibility or the problematic character of what happened (Komter, 1998). This occurs both in courtroom interaction (Drew, 1990; Ehrlich, 2001; Komter, 1994, 1998) and police interviews (Benneworth, 2006; Newbury & Johnson, 2006; Sliedrecht, 2013; Van Charldorp, 2011).

Witnesses and victims also do interactional defensive work to avoid blame attribution. An example is presented in Atkinson and Drew’s Order in Court (Atkinson & Drew, 1979). They studied the cross-examination of a police officer who testifies as a witness in a court about Catholic-Protestant violence in Northern Ireland. The witness uses defending components in different sequential environments: in response to explicit solicitation of accounts and accusations and in response to questions that do not explicitly accuse or solicit an account. In the latter case, the witness treats the question as part of a line of inquiry that is working toward blame attribution.

More recently, Fogarty (2010) examined accounts in Australian investigative interviews with child victims and witnesses (7–11 years old). She found that children talk about their agency in relation to crimes in three main ways: (a) children report their active resistance, (b) they account for why they could not stop the abuse from happening, and (c) they emphasize their lack of complicity (Fogarty, 2010, p. 283). She argues that children observably attend to the issue of who was agentive and that this suggests that they perceive agency as relevant for culpability. Consequently, children themselves do not treat the interview as simply retrieving uncontaminated information but as a social event in which issues of blame are at stake (Fogarty, 2010, pp. 307–309).

Similarly, MacLeod (2016) studied accounts by British rape victims in police interviews. They provide accounts that (a) counter the possibility that they were reckless; (b) display that they tried to avoid the use of alcohol and drugs but that the suspect was very persistent; (c) make clear that suspects were trustworthy, especially when they are known to the victim, and that the victim had no reason to be suspicious; (d) show that the victim offered appropriate resistance or provide a reason why they could not do so. These accounts are designed to preempt potential attempts to attribute blame to the victim (MacLeod, 2016).

Interactional research on the guidelines for interviewing vulnerable witnesses

As was explained, police officers who interview vulnerable witnesses (children and people with intellectual or cognitive impairments) should abide by a range of guidelines. Various CA studies of these interviews hightlight the importance of the guidelines in the ongoing interaction by showing if and how interviewers orient to these guidelines and how the interaction unfolds (Antaki et al., 2015a; Iversen, 2019).

First, some CA studies reveal that police officers do not always follow the guidelines. For example, the guidelines in England and Wales advise not to question victim’s behavior, yet police officers sometimes do exactly that (Antaki et al., 2015b). The same guidelines state that police officers should elicit demonstrations that vulnerable witnesses understand the difference between truth and lies. In the actual interviews, police officers regularly invite confirmations that truth and lies have been discussed (Richardson et al., 2018). The guidelines further suggest that the discussion of truth and lies should be brought up in the beginning of the interview only; the analysis shows that police officers also reiniated after the formal opening phase of the interview. This transforms the interview in something more like a suspect interview and risks communicating that the police officer does not believe the witness (Richardson et al., 2018).

Second, even if police officers adhere to guidelines, this sometimes achieves something else than the guidelines were written for. Childs and Walsh (2018) examined the recommended practice of asking the vulnerable witness toward the end of the interview whether there is something else they would like to say. This recommendation is motivated as soliciting more talk from the witness. Yet the fact that the question is usually produced in the closing phase of the interview makes it hearable as a formulaic step in pre-closing that prefers a no. This is indeed how it appears to work in the police interviews, especially when police officers frame the question as “before we can finish off then” (Childs & Walsh, 2018). Similarly, the truth and lies instruction is advised to understand the witness’s “understanding of the difference between truth and falsehood” and to emphasize the importance of telling the truth (Ministry of Justice in England and Wales, 2011: sections 1.2 and 8.4). However, inspection of actual interviews shows that police officers regularly require multiple demonstrations and confirmations of truth and lies. This treats the vulnerable witness’s turns as untruthful, rather than accomplishing instructing the witness. Additionally, reinitiations of the truth and lies issue later in the interview accomplishes a range of actions, such as implying that previous answers were lies (Richardson et al., 2018). Such discrepancies between what particular guidelines were written to achieve and what they actually achieve can probably be attributed to the conceptualization of language as an exchange of information (Fogarty, 2010). Fogarty (2010) shows that much research concerning investigative interviewing with vulnerable witnesses views language as a tool for exchanging information, rather than for social interaction. This is in line with CA literature that points out that much advice on interactional practice is regularly based on idealized ideas of talk (e.g., Sikveland et al., 2019).

Third, inspection of actual interviews with vulnerable witnesses recurrently reveals a tension between the guidelines to gather evidence that is usable in court (i.e., not the result of suggestion and detailed) and the guidelines to be supportive and build rapport. Antaki et al. (2015a) demonstrate that police officers usually do not acknowledge vulnerable witnesses’ displays of distress, nor do they affiliate with the witnesses’ stance regarding emotional trouble. The authors suggest that this lack of acknowledgment can be attributed to the guidelines—namely, that evidence should be admissible in court, a testimony that can be seen as being “solicited impartially and unjudgementally” (Antaki et al., 2015a, p. 429). Another study that highlights the tension between information gathering and maintaining rapport comes from Childs and Walsh (2017). They study police officers’ self-deprecating remarks (like “I’m going deaf, that’s all”) to warrant their requests for repair. With the self-deprecating remarks, police officers claim responsibility for the trouble leading to the repair request and thus avoid attributing blame to the witness. Consequently, police officers can be seen to do extra interactional work to elicit more detail while maintaining rapport. Yet another example comes from a study by Iversen (2019). She notes that investigative interviewing often includes detailed questions about issues that should go without explanation. That is, competent members of a speech community are supposed to know and understand without questioning why it feels safe to be with one’s mom or that a dad hitting and pushing a mom is frightening (Iversen, 2019, p. 15). Investigative interviewers presumably ask about such issues nevertheless to avoid assumptions and thus adhere to guidelines of evidence gathering without suggesting versions of what happened. This is, however, problematic for children, who orient to interviewers as not understanding them. Thus, there is a tension between the requirements of building rapport and being supportive on one hand and requirements of evidence gathering on the other (Iversen, 2019).

In this article, we further explore the tension between the attempts of neutral evidence gathering and doing what is relevant for the ongoing interaction in terms of approval or affiliation. We show how this dilemma is observable in actual interaction, both in uptakes by the police and in subtle pursuits by children that treat police officers’ neutral responses as insufficient. Hence, the article adds to existing literature that a police officer’s orientations to neutrality, as prescribed by the guidelines, cause interactional difficulties for both the police officer and the child.

Data and methods

The data consist of 30 video recordings of Dutch police interviews with child witnesses, adding up to 35 hours and 19 minutes of recorded materials. The recordings were prerecorded by the police as a part of the criminal procedure, for reasons including transparency, training, and assessment of reliability. They were obtained from two child-friendly interview rooms. The children in this data set are boys and girls between 6 and 11 years old, and they report some type of alleged sexual violence. One of the child-friendly interview rooms only provided the year of birth, hence, for some children two possible ages are mentioned.

The recordings used for this analysis were obtained from the police with permission of the Public Prosecutor’s office. The recordings were stored in a safe, could not leave the university building, and were only shown to researchers within the project. Moreover, all recognizable information in transcripts was replaced by pseudonyms, full transcripts cannot be published, and publications and the Public Prosecutor’s office screens publications for information that could lead to indentification. However, due to legal restrictions, we as researchers were not allowed to ask for informed consent for research purposes retrospectively. We explored various possibilities to seek informed consent, but all of these options came with ethical and other problems (Jol & Stommel, 2016a). Finally, we argued that the recordings were created with informed consent, that the recordings are more appropriately seen as archival data, that the research including its knowledge utilization purposes was not in conflict with the original purposes of recording, and that most ethical guidelines make qualified exceptions for the informed consent requirement in case of archival data.The faculty’s Ethical Research Committee approved of the project.

The data were transcribed using Jeffersonian conventions. We used the conventions for embodied behavior developed by Mondada, 2018), and we included as much embodied behavior as was relevant for the analysis, indicated by “p” and % for the police officer and “k” and * for the child. The translations stay as close to the Dutch original as possible. Parts of speech that have to be added to the translation in order to be comprehensible are marked between // and //.

Interviews with child witnesses in the Netherlands take place in special child-friendly interview rooms that are furnished to look cozier than regular police interview rooms and that are equipped with cameras and a microphone. Police officers need to be certified, and they work in pairs: One police officer does the actual interview, the other operates the equipment and coaches the interviewing police officer.

Interviews usually include the following phases, roughly in the following order:

• an introduction phase,

• giving instructions or ground rules (e.g., that the child should correct the police officer when necessary),

• a free narrative or free recall phase (when the child does the talking and the police officer listens),

• a questioning phase (when the police officer elicits more details about the story)

• a “break” or directions consultation (in which the interviewing police officer usually leaves the room and consults the coaching police officer, e.g., about which questions should still be asked)

• a closure phase (when the police officer thanks the child and gives the child the opportunity to ask questions; but see Childs & Walsh, 2018).

In previous analyses (Jol, 2020; Jol & Stommel, 2016b), it struck us that children regularly provide accounts of their own role without being invited to do so and that police officers reacted rather reticently. The transcripts were searched for accounts of the child’s own behavior that were not solicited by the police officer. Such reports occurred in the following sequential environments:

1. In narratives—for example, in the free narrative phase. These turns may be invited by the police officer with questions like “What is it that you have come to tell me?” The invitation often presumes that there is something relevant to tell (Kidwell, 2009), but the question does not necessarily elicit an account of the child’s behavior.

2. When the child answers more than the question (cf. Stivers & Heritage, 2001). That is, when the child provides an answer but elaborates that answer more than what was necessary given the question.

3. After the answer uptake. For example, after a sequence-closing third-position “okay,” the child presents an account of his or her own role.

We included both accounts that are usually indicated as accounts of (i.e., describing the situation in an understandable way) and accounts for (i.e., justifications, warrants, explanations, excuses). These two are related as accounts for become relevant when behavior fails to make sense straightaway (Buttny & Morris, 2001, p. 287). Moreover, there is no clear difference in design: Descriptions of can do defensive and justifying work (Atkinson & Drew, 1979; Potter, 1996). They have in common that they make relevant the child’s behavior, without being prompted.

In total, we found 95 instances of unsolicited reports of the child’s role in the incident. There was no relationship between the length of the interview and the occurrence of accounts. In eight interviews, no instances were found. Of 22 interviews with instances, four interviews provided 34 unsolicited accounts. These are the interviews by four different police officers with Myrthe (10 years old; 14 sequences), Dorien (10 or 11 years old; nine sequences), Lieke (9 or 10 years old; 11 sequences), and Jenna (8 or 9 years old; 10 sequences). In 18 interviews, one to six unsolicited reports were found. Some children clearly provided more unsolicited reports of their own behavior than others. The children who produced most of the instances were girls (cf. Stivers, 2012), mostly in the upper age range from the children in the corpus. At the same time, the unsolicited reports of their own role are certainly not unique to these four children, nor to these girls. All instances were analyzed using CA.

Analysis

We first discuss how children construct accounts of their own behavior. This part of the analysis is organized along the continuum of more factually presented accounts to accounts with overt normative orientations. We discuss children’s unsolicited accounts of their own role, how this includes reporting resistance and how children sometimes display an normative orientation to resisting the offender which invites an affiliative uptake. Then, we examine how police officers respond in ways that are designedly neutralistic, thus not producing affiliative uptakes. Finally, we discuss an example of a child who pursues something other or more than neutralistic uptakes.

Children’s accounts: Reports constructed as factual

Children regularly provide factual reports about what happened. At the same time, they can also make it clear that what happened was not their initiative or that they did not enjoy it. Three ways to do that are illustrated in the following excerpt from the free recall phase in an interview with 8-year-old Wencke about a man in the park.

EXCERPT 1: and then suddenly a gentleman came

Let us first consider the turns in lines 3–9. Wencke uses the construction “we were doing X, when Y happened” (Kidwell, 2009; Wooffitt, 1992) as an introduction³ to a sequence of reported events marked by “and then” (lines 12, 16, 20). She begins sketching a fairly ordinary scene: She went playing with a girlfriend in a place nearby her house (lines 3–8). In line 9, she introduces something new: “and then suddenly a gentleman came.” The use of passivity marker (Komter, 1994) “suddenly” is notable here. It constructs the arrival of the person as something unexpected and unanticipated. This construction thus works to portray Wencke and her friend as not having been looking for the suspect, not seeking trouble, and definitely not initiating that the man should approach them. It also implicitly counters a potential blame-attributing version of the story that they could and should have left earlier: Because they did not expect him to approach them, they had no reason to leave earlier.

The second point to be made about Excerpt 1 concerns line 12. It is an example of what Ehrlich called “the grammar of non-agency” (Ehrlich, 2001). In line 12, the girl reports that the suspect pulled her on his lap, rather than for example, “I sat down on his lap” or “I had to sit on his lap.” Both alternatives would have indicated purposeful action from the child. Hence, the linguistic construction presents the suspect as the acting person here and the girl as the passive object of his actions (e.g., Deckert, 2010; Ehrlich, 2001; Pomerantz, 1978).

Finally, in line 14, the girl explicitly takes a negative emotional stance toward the man’s actions with “I didn’t want that.” This practice is recurrent in the data and makes explicit that what the man was doing was against her will from the beginning. By doing so, Wencke both reports the man’s actions as problematic and her own role as involuntary. It is also relevant to consider when she produces “I didn’t want that.” In the previous line, she sets off to continue reporting the course of events with “and then •hh –, ” which typically would be finished by a phrase like “I/he did/said/saw Z.” Yet she does a self-repair and begins a new turn constructional unit “I didn’t want that,” thus showing an orientation to this utterance as important to produce here. The girl’s involuntariness and the problematic character of the man’s behavior are further built up in lines 17–19: when she wanted to leave, but the man was still holding her.

The excerpt shows several ways in which children provide an account of their own role as not wanting the abuse to happen, having innocent intentions, not doing anything to invite the abuse, not having seen the trouble coming, and therefore having had no reason to take action. Rather than presenting themselves as actively taking a role to accomplish something in the story, these features allow the children to assume a role as rather passive participants with no bad intentions and to whom the abuse happened. They are “doing being the victim.”

Children’s accounts: Reporting resistance and normative orientations to resistance

Children also frequently (in 53 instances in total) highlight that they resisted the suspect—that is, that they protected themselves to (further) (potential) harm—and how well they did that. Excerpt 2 illustrates this. The sequence occurs almost 36 minutes into the recording of an interview with 9-year-old Delphine about repeated sexual abuse by her father. The lines of interest are lines 17–20, in which the girl reports how she escaped after the abuse by misleading the offender:

EXCERPT 2: I really pret(h)ended that I was falling asleep

In line 1, the police officer poses the question “how it stopped”; that question is answered in lines 2–9. The police officer receives this answer with “okay” (line 10) and an answer-repetition (lines 11–12). Delphine confirms the repetition with “yes” (line 13). The police officer projects a next sequence with ‘”mkay” (line 15) (Beach, 1993; Gaines, 2011) and directing her gaze to her notes. The girl, however, expands her answer with a rather dramatic move before the police officer can produce a next question. The girl’s initial answer in lines 2–9 reports the ending of the abuse as something she did not pay particular attention to, by using “dunno” (Potter, 1996), and as nothing special, using “just” (gewoon). By contrast, her elaboration (lines 17–21) portrays the ending of the abuse as the result of fooling the suspect by pretending to fall asleep. This, in turn, allowed her to escape and go to her brother Berend. Hence, this second version of how the abuse ended highlights her agency.

The excerpt also illustrates one of the ways children take a normative stance toward resistance toward the alledged offender. We found that children not only repeatedly report such resistance, they also display a normative stance. The first way that we identified is laughter. In the previous excerpt, the laughter comes in precisely when the girl accounts having successfully outsmarted the suspect (line 17). She thus presents her resistance in a cheerful way. This makes it possible to hear the report as proud and to hear the report as having an implicitly normative orientation to resistance as the right thing to do (cf. Jefferson, 1984). We found this way of taking a normative stance toward resistance in other interviews as well (12 sequences from four interviews).

A second way of taking a normative stance toward the norm of resistance are children’s accounts for not providing resistance during or after the incident (e.g., “I couldn’t do anything else,” 6-year-old Ben, Excerpt 4), not leaving (12 cases), or not reporting the abuse straightaway “because I was a little afraid of the consequences,” 8- or 9-year-old Jenna). Komter’s study of knowledge in Dutch courtrooms provides an interesting example of how a deviation of such norms can lead to damaging inferences (Komter, 1995). She gives an example of a defense lawyer who explicitly stated that not reporting rape immediately negatively affected the reliability of the victim. Children’s accounts for not telling someone straightaway and not resisting anticipate such potential reproaches and deal with them by making clear why they did not resist or escape the suspect (cf. MacLeod, 2016).

The third and most salient displays of normative orientations to resistance occur in self-evaluations. Self-evaluations can be either positive evaluations of resistance or negative self-judgments. Positive evaluations are quite rare (four sequences in three interviews), for example, an interview with a boy named Teun. He reports having said “no” to an unknown man who wanted him to touch the man’s private parts. He then adds “•hh and my mother was proud of that?” This implies that the child’s behavior is under evaluation and that, apparently, there are courses of action to be proud of. Although these reports are uncommon, they suggest that resisting is a relevant norm for victims of sexual violence (see also Excerpt 4). Negative self-evaluations or self-judgments occur slightly more often (eight sequences in four interviews). In Excerpt 3, 9- or 10-year-old Lieke judges herself for answering the suspect’s question where she lived and not being more resistant. The suspect is an unknown man at an open air festival who later pushes her over and tries to put his hand in her panties.

EXCERPT 3: and so that was very-also very stupid

The self-judgment starts in line 2. Lieke portrays answering the man’s question (line 3) with the idiomatic expression “like chicken without a head” (als een kip zonder kop). Although this is not yet formulated as an explicit evaluation, she displays a retrospective understanding of her own action as not properly thought through. In this case, the accompanying laughter supports her presentation as having been silly. The police officer treats the girl’s narrative as not yet finished by producing a continuer (line 4). Next, the girl provides an explicit self-judgment (line 5). She judges her own behavior negatively by deeming answering the suspect’s question as “very stupid,” emphasizing this even more by stretching the sound. The self-judgment invokes the norm that “one should not tell strangers where one lives.” With this self-judgment the girl suggests that she would have done things differently in retrospect. This is further supported by claiming “I didn’t know anything” (line 7), which suggests that she knows now and which thus excuses her.

Interactionally, the self-judgment is a stake-confession. Potter (1996) points out that speakers sometimes make explicit that their account could be heard as merely informed by their stake, for example, when politicians praise their own political party. A prototypical way of formulating a stake confession is “But I would say that, wouldn’t I?” (Potter, 1996, p. 130). Potter suggests that the stake confession may work as a display of honesty and objectivity. Moreover, it makes it more difficult to undermine such a statement interactionally because an objection of self-interest has become redundant. The girl classifies her action as accountable and stupid and she provides this evaluation herself. This self-judgment makes a judgment by the police officer redundant and thus renders the child’s report less susceptible to judgments by the police officer and a potential overhearing audience. Even stronger, the preferred response to self-deprecations is disagreement (Pomerantz, 1984). This indicates that after the child’s judgment, a denial by the police is relevant. The police officer refrains from providing an uptake in line 6 and thus does not align with the girl’s action. Only after the justification (line 7), he provides a minimal uptake (line 9) (we elaborate on this uptake later).

How police officers respond: Neutralistic uptakes

This section moves from children’s accounts to police officers’ responses to such volunteered reports of the child’s behavior. It is important to note that an uptake is not always relevant after children highlighted, for example, how they resisted or account for why they did not resist. When children continue their talk straightaway, there is no slot for an uptake by the police officer. In these cases, the report of the child’s behavior is constructed as a discourse unit, and it would take interactional effort from the police officer to “jump” back (cf. Sacks, 1987, p. 58).

However, the police officer is regularly provided with a slot that, in many other contexts, could appropriately be filled with an affiliative uptake. This section shows that police officers generally avoid producing such uptakes. Chevalier and Moore (2015) call such avoiding practices “restricted activities.” This term indicates that parties in different types of institutional interaction withhold particular activities, such as affiliating and assessing. By withholding these activities, they contribute to a particular type of institutional context. Excerpt 4 from the free recall phase of an interview with 6-year-old Ben provides an example. It is part of a narrative about his older cousin who proposed to rub together naked body parts, including private parts. The police officer is writing throughout the excerpt.

EXCERPT 4: I couldn’t do anything else

Ben clearly orients to his own participation in the activity as accountable and justifies it by presenting participating as the only option. The police officer had the opportunity to respond in line 8, for example, by acknowledging the distress displayed by the tearful voice (transcribed by combining the symbols for creaky voice # and wobbly voice ~) (e.g., “poor thing” or “a:::h”) or reassuring Ben that it was not his fault. Instead, she continues to make notes without verbal comments.

The avoidance of committing to the child’s report is characteristic of police officers’ uptakes (Antaki et al., 2015a). An example can also be found in Excerpt 2 where the police officer’s uptake is a brief silence and “okay” (line 23). Other uptakes in the data include repeating a version of what the child said; next questions or clarifying questions; “hmhm,” “yes,” or combinations of such responses.

This broad range of uptakes (or lack thereof) have in common that they do not overtly approve of the child’s behavior or reassure children that it is not their fault. There are prosodic variations that make some uptakes more affiliative than others, but the uptakes are very minimal and do not take an overt stance toward the child’s reported behavior. The uptakes accomplish aligning rather than affiliating (Antaki et al., 2015a). In short, these uptakes are “neutralistic” (Clayman & Heritage, 2002, p. 120). This term highlights that neutral communication is an idealization because speakers always do something and assume a role when they say something (cf. Robinson & Sheehan, 1983). The term neutralistic indicates that interlocutors display “being neutral” rather than achieving actual neutrality.

Neutralistic uptakes are not necessarily unsuitable. When children construct their report of their own behavior as more factual, this generally does not make relevant reassurance or approval, and therefore uptakes like “hmhm,” “okay,” or a next question are interactionally appropriate. The police officer’s lack of stance becomes problematic when the child’s normative orientation becomes more overt—for example, when the child displays being proud or provides an explicit justification or evaluation—and when the child leaves a slot for the police officer to respond. For example, in Excerpt 5, Myrthe explicitly evaluates her own choice not to comply with the suspect’s request to show her belly via Skype (line 9).

EXCERPT 5: I’m very glad about that

Myrthe does several things in this excerpt, including creating a contrast between her friend Pien as compliant with the suspect’s request and herself, who resisted the suspect. The latter is produced with a smile and laughter (line 8) and is interpretable as being proud. She also explicitly assesses her own behavior in a positive way (line 9) and supports that assessment with laughter (line 10). This assessment makes relevant an agreement or approval of the girl’s account of resistance. However, the police officer ignores the child’s invitation to agree and proceeds with a factual next question about what was said in lines 1–6 that elicits more detail about what exactly Pien showed. The police officer thus stays clear of taking sides. Simultaneously, the notable lack of uptake makes available the inference that the police officer does not agree with the child’s assessment. Another consequence of the lack of uptake is that the norm of resistance is treated as unproblematic, whereas such norms can be used as a resource of victim-blaming. In effect, not probing or contradicting that norm implicitly confirms the norm as not needing explanation (MacLeod, 2016).

The police officer’s dilemma: Neutralistic uptakes or doing what is interactionally relevant?

The police officer’s dilemma between “being neutral” and doing what is interactionally relevant surfaces in the interaction. Police officers only rarely produce more than a neutralistic uptake, but even in those cases, the demonstration of affiliation is very minimal. We already saw Excerpt 3, partially repeated here as 3a:

EXCERPT 3a: and so that was very-also very stupid

The police officer acknowledges the child’s self-judgment and her account with a minimal “hm,” a shrug, and headshake. This embodied behavior is understandable as responsive to the child’s account as troublesome and⁴ doing something like “don’t worry” and thus doing reassuring. It also works as an encouragement to continue the story. All this work is very implicit, minimal, and hardly on record. It can be seen as a trade-off of incongruent orientations to being neutral and displaying what is interactionally relevant.

Excerpt 6 offers another example of a trade-off between “doing being neutral” and affiliation. The excerpt is taken from the questioning phase of the interview with 7-year-old Karin. She reports that she was playing outside when an unknown man asked her if she wanted candy, invited her into his home, and asked her to lick his private parts and to show him her private parts. Directly prior to the excerpt, the police officer asked what she did when the man asked her to show her private parts, to which she said “no” because she is not allowed to do that.

EXCERPT 6: yes; well that is an idea

Karin volunteers a strategy about what she would do in the hypothetical situation that someone asks her to show her private parts (lines 3–6, 8–16). She would try to put that person off by suggesting that her private parts are both difficult to reach because her clothes are very tight and that they are very dirty (cf. Woodhams et al., 2012). She thus portrays herself as having thought about such a situation, being smart and being prepared. The police officer receives this⁵ plan as new information using “oh” (line 18) (Heritage, 1984), rather than, for example, “that’s a smart plan.” The girl treats this “oh” as a sign of potential misunderstanding by explicating that she would say this as an excuse (line 19). This stresses the nonfactual character of the “flies-claim” and thus orients to the possibility that the police officer has misunderstood her strategy as factual.⁶

Now that the girl has unambiguously reported that she prepared a strategy, approval is relevant. The police officer claims understanding now (line 20), in slight overlap with “yeah” (line 21), and then continues with “well” (Mazeland, 2016). She acknowledges that the child’s proposed strategy is an idea, but she notably does not use an adjective like “clever” or “good.” This utterance could have been a downright disapproval. Yet the rising intonation makes it ambiguous between a less than enthusiastic assessment and a slightly positive acknowledgment of the idea. Furthermore, the lexical choices allow the police officer to account for her response in a literal manner by saying that she only said that the plan is “an idea.” This can be seen as not evaluative or adding information and thus as neutralistic. Thus, it is observable that the police officer does interactional work to find middle ground between approval and “doing being neutral.”

Responding to the uptake: Children’s pursuits of more than a neutralistic uptake

The question from a conversation analytical perspective is, of course, whether children orient to the neutralistic uptakes as insufficient or inappropriate. In our data children generally attend to the next question, as Myrthe does in Excerpt 5, or continue their narratives or answers, as Ben does in Excerpt 4. Hence, children generally do not immediately treat neutralistic uptakes as problematic within the interaction.

However, children sometimes pursue a more explicitly aligning or affiliative uptake. Most of the relevant sequences come from two interviews that have already been introduced: the interview with Lieke about the man at the festival and the one with Myrthe about the Skype session. It occurs in some other interviews as well, though even less frequently (for example, from another interview, see Jol et al., 2019, pp. 121–124). Children do not pursue more affiliative uptakes by explicit requests. Rather, they provide response cues, repeat or rephrase their accounts for not resisting (enough), provide an explicit evaluation, or emphasize how successful their resistance was. The following excerpt is an example of such a pursuit. It is taken from the interview with Lieke about the man at the festival, 13 minutes into the recording, in the questioning phase. The police officer invites her to narrate what happened next (line 1) and specifies from which point the girl should resume the talking (lines 2–4).

EXCERPT 7: and then I started shouting no and kicking him

The girl resumes the narrative as requested by reporting that the man tried to get under her shorts (line 5) and by explaining how the man could do so by explicating that she was wearing shorts (line 6). Then she shifts to her own role and claims that she now understood what the man was up to with a somewhat apologetic smile (line 7), possibly treating it as something that she should have understood earlier. She then reports running away juxtaposed to “got it,” suggesting that she ran away immediately (line 9) with a smile that makes it hearable as a proud report. The report works to depict herself as potentially blaming herself for not grasping the man’s intentions earlier but also as taking action as soon as she knew what the suspect was after. The word choice “running” portrays her action as urgent.

The police officer receives the account with “okay” (line 10) and a request for more detail about how the suspect tried to get into her pants (line 11). The pitch in “okay” is higher and may be heard as appreciative. Yet it is clearly not overtly approving of the child’s resistance. The girl provides an answer to the police officer’s request for more detail by explaining that and demonstrating how the man put his hand in her trouser leg (lines 13–15). The police officer provides another acknowledgment “okay” (line 17), using rising intonation that presents the uptake as surprised or as a news receipt.

At this point, Lieke again moves on to reporting her own resistance, but in an upgraded and expanded version. She begins reporting “and then I ran,” which would bring her to the end of the narrative, but she interrupts herself. Rather than reporting “merely” understanding and running away, she now enacts shouting “no” (though presumably not at full volume) and kicking the man (line 20). She also portrays her resistance as effective by reporting that the man got off her (line 23; see also Excerpt 2). Finally, she upgrades “running away” (line 9) to “running away very hard” (line 24). These upgrades emphasize her resistance and hence arguably pursue an uptake that is more affiliative than the “okay+question” (lines 10–12) (cf. Weatherall & Stubbe, 2015), e.g., approval (“good that you did that”). The police officer picks neither option and proceeds by asking for more detail (lines 25–27).

The pursuit of affiliation (reassurance, approval) does not only occur in this interview in the aforementioned sequence. Already in the free recall phase, the girl mentions that the man “luckily didn’t succeed” and that she ran away. Roughly 10 minutes into the interview, the girl also brings up shouting “no,” kicking, and running away. She returns to her own resistance without solicitation, adding up to 14 times (including pursuits).

Such moves invite the police officer to do more than doing being neutral and treat the neutralistic uptakes as insufficient. Hence, neutralistic uptakes can be interactionally problematic.

Conclusion and discussion

The analysis first showed that children often use unsolicited reports to display that they did not want the abuse to happen, that they were not reckless, that they did not see the trouble coming, and that they had innocent intentions. Such reports accomplish “doing being the victim.” Moreover, the analysis showed that children report that they resisted the suspect or tried to, either by being smart or by offering verbal or physical resistance. Second, we found that children orient to resistance in a normative way, by highlighting the resistance, accounting for their role, and sometimes providing evaluations, including self-judgments that deem their own role as not resistant enough. Children thus treat providing resistance as a normative requirement. They do so as early as in the free recall phase. This suggests that accounting is not something they learned in the course of the interview. The norms that children observably orient toward remind of some elements of rape myths that establish what a “real victim” is, specifically the requirement that a real victim should provide appropriate resistance. Children thus appear to use the unsolicited accounts to do “being a real victim” while anticipating and countering potential victim-blaming moves that would undermine their victimhood (Jol et al., 2019). These issues were also addressed in Fogarty’s study (Fogarty, 2010; MacLeod, 2016), but the focus on unsolicited accounts offers firm evidence that children themselves treat their own behavior as relevant for the interview.

Another finding is that police officers respond to such accounts in a neutralistic way. They avoid taking a stance toward the child’s accounts, and they generally acknowledge children’s accounts as information by providing uptakes such as “okay,” “hmhm,” pauses, answer-repetitions, next questions, and combinations of these. By doing so, they talk into being the institutional constraints that police officers have to comply with. Police officers are supposed to refrain from committing to the child’s story, since it is still under investigation. Accepting the child’s story too easily may be seen as disturbing the police’s primary goal of truth finding. The fact that the interview is not only monitored in real time by colleagues but also recorded and potentially scrutinized in the future during the criminal procedure adds to the importance of exhibiting conduct that can be seen as neutral. The police officer’s mostly neutralistic uptakes are in line with previous research into displays of emotions and distress in institutional settings where institutional neutrality is important (Antaki et al., 2015a; Hepburn & Potter, 2007; Weatherall & Stubbe, 2015) and confirm earlier research that signals a tension between neutral evidence gathering and rapport building (Antaki et al., 2015a; Childs & Walsh, 2017; Iversen, 2019).

The requirement of maintaining “neutrality” versus building rapport creates a dilemma. Responses like “okay” and “hmhm” might seem neutral in isolation or when the interview is regarded as exchanging information. Yet from an interactional perspective, they sometimes turn out to accomplish disaffiliation in ongoing interaction, affirm normative sources for victim-blaming procedures, and enable problematic inferences. The dilemma manifests itself in the interaction. When police officers produce uptakes that are slightly affiliative, the affiliative character tends to be minimal and implicit. They thus show a trade-off between interactional requirements and the constraint that police officers should be neutral. Second, children sometimes pursue a more affiliative uptake, thus displaying that something was lacking. Not all children do this, and pursuits can be quite subtle, but their occurrence nevertheless shows that neutralistic uptakes can be problematic for children (cf. Weatherall & Stubbe, 2015). Consequently, attempts to be neutral undermine other institutional tasks such as creating a safe environment for the child to tell his/her story. Police officers also confirm this dilemma. Jakobsen et al. (2017) asked Norwegian police officers to comment on their own interview with child witnesses, especially in sequences where children were clearly upset. The police officers frequently commented that they should have provided more support instead of being so focused on truth finding.

On a more general level, the analysis adds to previous studies that argue against the conduit metaphor of interaction (Hutchby, 2005, 2007; Phillips, 1999). The conduit metaphor (Lakoff & Johnson, 1980; Reddy, 1979) conceptualizes language (words, sentences) as containers of meaning. Those containers transfer ideas (feelings, thoughts, memories) from a sender to the receiver of the message. This metaphor is often implicit in many English expressions about language and interaction, such as “putting thoughts into words” and “giving an idea” and “hollow words.”

The conduit metaphor supports the idea that language use can be neutral. A speaker or writer is merely sending ideas to be unpacked by the receiver, and language thus can be a neutral way of transporting ideas. This way of conceptualizing language has found its way into advisory texts. As we saw in the introduction, the Manual for police interviews with children encourages police officers to interview children “in the most neutral way possible,” thus building on the idea that interaction can be mostly neutral. This is in line with manuals on mediation (Phillips, 1999) and therapy (Hutchby, 2005). As Hutchby summarizes, the professional can be seen as a “conduit, a largely neutral presence ‘drawing together’ and ‘re-expressing’ or clarifying” (Hutchby, 2005, p. 308).

The notion of neutrality has been challenged by CA research. Research has shown that contributions that are considered neutral in manuals (e.g., summarizing) fulfill decidedly institutional goals and steer and shape the interaction (Hutchby, 2005, 2007; Phillips, 1999). The conduit metaphor has thus been shown to hide particular aspects of interaction—namely, the institutional, goal-oriented aspects of the professionals’ contributions to the interaction. This article confirms that neutrality in interaction is not possible. More importantly, the analysis adds to this debate that attempts to achieve neutrality may be interactionally problematic.

Children’s orientations to resistance as the right thing to do also point to another dilemma for police officers (MacLeod, 2016). In the present analysis, police officers do not challenge or deny resistance norms. If police officers provided more affiliative responses (e.g., compliments or approval of resistance), they would implicitly confirm the norm that children should resist offenders. Hence, more overt affiliation in response to reports of resistance would be a solution to one problem but create a new dilemma. Police interviews are, hence, a very complex, layered setting that involves various social and professional norms that may collide in various ways.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The research was funded by the Centre for Language Studies (CLS) of the Radboud University Nijmegen.

Notes

¹ Some ethnographic work accompanied was done to better understand the setting. This is more elaborately explained in Jol (2020).

² This article is based on an earlier version of the analysis in Jol’s dissertation (Jol, 2020).

³ This phrase is a bit odd in Dutch. A more standard phrase is, e.g., “next to us/our house stands a flat.”

⁴ The Dutch utterance is missing a subject.

⁵ Hij is translated here as “it” because it likely refers to her underwear or pants.

⁶ Diminutives are very common in Dutch. It is uncertain if the diminutive broekje is just one of those cases, if it refers to children’s pants or to shorts. The latter seems more likely given the embodied demonstration. Hence the translation as “shorts.