Did Your Mom Help You Remember?: An Examination of Attorneys’ Subtle Questioning About Suggestive Influence to Children Testifying About Child Sexual Abuse

Abstract

Researchers studying children’s reports of sexual abuse have focused on how questioners overtly assess coaching and truthfulness (e.g., “Did someone tell you what to say?”). Yet attorneys, and defense attorneys, in particular, may be motivated to ask about suggestive influence and truthfulness in subtle ways, such as with implied meaning (e.g., “Did your mom help you remember?”). Such questions may be particularly challenging for children, who may interpret statements literally, misunderstanding the suggested meaning. The purpose of this study was to examine and categorize how attorneys’ ask about suggestive influence and truthfulness. We wanted to learn how attorneys subtly accuse suggestive influence, and how frequently this occurred. We hypothesized that questions indirectly accusing suggestive influence would be common, and that defense attorneys would ask more subtle questions, and fewer overt questions, than prosecutors. We examined 7,103 lines of questioning asked by prosecutors and defense attorneys to 64 children testifying about alleged child sexual abuse. We found that 9% of all attorneys’ lines of questioning asked about suggestive influence or truthfulness. The majority (66%) of these were indirect accusations. Indirect accusations of suggestive influence spanned a range of subtleties and topics, including addressing conversational influences (e.g., coaching), incidental influences (e.g., witnessing abuse), and others. We also found defense attorneys were less likely than prosecutors to ask about suggestive influence and truthfulness overtly. We conclude that attorneys commonly ask about suggestive influence and truthfulness in subtle ways that developing children may struggle to understand, and which may result in affirmations of influence, even when allegations are true.

In criminal cases of alleged child sexual abuse (CSA), children’s credibility often hinges on their testimony. As children are susceptible to suggestions, both from those who aim to influence their reports and those who aim to establish what children are alleging, how attorneys question children is of great importance. Unsurprisingly, prosecutors and defense attorneys dedicate a substantial amount of questioning to establishing and attacking, respectively, children’s credibility, including by asking about coaching and suggestive influence (Denne et al., 2020). But unlike other jurisdictions where attorneys must explicitly state the position that the child’s report is false or the product of influence (Hanna et al., 2012), in the United States, attorneys can ask about truthfulness and suggestive influence subtly (Myers, 1986, 1994). For example, rather than asking overtly if the child lied or was coached (e.g., “Did someone tell you what to say?”), attorneys in the United States can subtly imply that the child’s report was false or influenced (e.g., “Did you talk with your mom about what you would say today?”).

Allowing attorneys to accuse lying and suggestive influence subtly is a problem. Children may lack the cognitive skills needed to understand implied meaning and motives, such as the theory of mind, which is the ability to recognize that others’ beliefs about the world can diverge from their own and to attribute false beliefs to others (Premack & Woodruff, 1978). Defense attorneys may be motivated to exploit children’s cognitive deficiencies by using subtle questions to imply children were coached or influenced. If children testifying fail to understand the implied meaning of such questions, they may acquiesce to these credibility attacks, unintentionally admitting coaching or suggestive techniques, even if such suggestive influence did not occur prior to cross-examination. Furthermore, to the extent that prosecutors also use subtle questioning styles, they may unintentionally exacerbate, rather than mitigate, credibility concerns.

Beyond what attorneys ask about, the phrasing of attorneys’ questions to children—whether questions are subtle or overt—therefore, can resolve or produce concerns about the credibility of accusations, including concerns about influence and truthfulness. Given the potential consequences, understanding how attorneys question children about suggestive influence is important. Yet prior researchers have focused on overt questions. We are not aware of former investigations that systematically examine the frequency of subtle questions about suggestive influence. Nor are we aware of any investigations that examine what these subtle questions look like. This was the purpose of this study. Furthermore, we wanted to assess the range of questioning practices used when attorneys inquire about suggestive influence.

Indirect Attacks on Credibility: Polysemous Implicatures

One way that attorneys may subtly suggest children’s reports are the product of influence is to ask polysemous implicatures. “Polysemous” refers to the numerous understandings of statements (Klein & Murphy, 2001), and an “implicature” is a statement that implies meaning beyond the literal sense (Grice, 1975). A “polysemous implicature,” refers to a phrase or utterance whose implied meaning differs from the explicit meaning (Noveck, 2001). Individuals’ understanding of implied meaning depends on linguistic and cognitive skills that develop with age, such as the ability to understand that others’ intentions and beliefs may differ from one’s own (Premack & Woodruff, 1978), and the ability to recognize that words have multiple meanings (Bloom, 2002). As these skills continue to develop into adolescence, children may struggle to understand polysemous implicatures (Sullivan et al., 1994; Wimmer & Perner, 1983).

While some researchers have addressed children’s ability to comprehend conveyed meaning (Beal & Flavell, 1984; Bonitatibus, 1988; Clark & Lucy, 1975), we know of only one study that has examined polysemous implicatures in the context of coaching (Wylie et al., 2020). In an examination of 5–10-year-old acquiescence to implied (e.g., “Helped you remember what to say,”) and explicit (e.g., “Told you what to say”) coaching questions about a hypothetical disclosure of an adult’s transgression, Wylie et al. (2020) found that children acquiesced to questions subtly implying coaching even when no coaching occurred. The subtler the question, the more children acquiesced. These findings indicate that subtle questioning in the form of polysemous implicatures may be particularly problematic for children testifying in court. However, it is unclear if the polysemous implicatures examined by Wylie et al. are akin to real investigations. While it is plausible that defense attorneys use polysemous implicatures in cross-examination to subtly suggest influence resulting from conversations with others, other unidentified ways of implying suggestive influence may also be used. The purpose of this study was to examine and categorize these subtle attacks.

Courtroom Assessments of Suggestive Influence

The courtroom is especially relevant for assessing children’s understanding of polysemous implicatures, because children’s limited ability to infer the implied message of questioning may lead their credibility to be susceptible to subtle attacks. Specifically, attorneys are instructed to ask the child to specify the adults involved in questioning the child (Bailey & Rothblatt, 1971), and whether adults “helped them remember,” or “practiced” what might happen in testimony (Myers, 1998). Children are unlikely to recognize that the attorney could be using these phrases to imply coaching (Lyon & Stolzenberg, 2015). Children who struggle to understand attorneys’ intended meanings behind such subtle questioning may unintentionally make statements suggesting their reports were fabricated or influenced, even when coaching or other suggestive techniques have not been used. Even if children possess a basic understanding of polysemous implicatures, the courtroom context may diminish this. Specifically, it is very likely that performance on these questions would be harmed by issues related to the unique set of cognitive and emotional challenges inherent when providing legal testimony.

While ample amounts of laboratory and field research on forensic interviewing exist, only a select number of studies have examined how children are questioned in court (Andrews et al., 2015a; Andrews et al., 2015b; Evans et al., 2009; Evans & Lyon, 2012; Klemfuss et al., 2014; Stolzenberg & Lyon, 2014, 2017; Zajac & Cannan, 2009; Zajac et al., 2003). In a study of 5–12-year-olds testifying in court, Stolzenberg and Lyon (2014) found that defense attorneys asked overtly if accusations were coached or fabricated in only 21% of cases; prosecutors did so in only 26%. In contrast, in another examination of trial transcripts of 5–17-year-olds’ testimonies, Denne et al. (2020) found that 21% of attorneys’ credibility-related questions asked about suggestive influence and honesty (with defense attorneys asking more questions about suggestibility and honesty than prosecutors; 29% compared to 18%). Yet, Denne et al. (2020) did not examine what these questions looked like in terms of how many were subtle.

The infrequency of overt coaching and honesty questions noted by Stolzenberg and Lyon (2014) suggests that attorneys are likely asking about these topics in subtler ways, such as with polysemous implicatures. For example, these researchers found that defense attorneys were more likely than prosecutors to ask about specific disclosure recipients, about more recipients, and about the content of conversations with different recipients (Stolzenberg & Lyon, 2014). They noted defense attorneys may use these questions to impeach children’s credibility by highlighting inconsistencies across disclosure conversations (Stolzenberg & Lyon, 2014). Defense attorneys may also use questions about prior conversations to subtly suggest coaching or suggestive influence by highlighting opportunities to be influenced (Lyon & Stolzenberg, 2015). However, examining subtle attacks on credibility was outside the scope of their study, so how attorneys subtly ask about coaching and suggestive influence remains unexplored.

This Study

While researchers have examined how attorneys question children about suggestive influence, they have focused on questions that do so overtly. We examined how attorneys asked children about prior suggestive influence and truthfulness, focusing on subtle questioning, including polysemous implicatures. Doing so is important, as children may be vulnerable to these subtle questions, potentially exacerbating miscommunications. Understanding how attorneys address suggestibility and truthfulness in court can help to develop procedures that better distinguish true from false allegations—procedures that minimize children’s vulnerabilities and maximize competencies without infringing on the adversarial process.

We began with a few exploratory hypotheses. First, we expected that questions about suggestive influence would occur often. While we know that attorneys in the U.S. question children—at least to some extent—overtly, they are likely also addressing these issues in subtle ways not previously detected. Second, given defense attorneys may be motivated to strategically induce admissions of coaching through subtle questioning, we expected that defense attorneys would ask more subtle questions about suggestive influence, while prosecutors would ask more overt questions. Our primary goal was to examine what questions subtly accusing suggestive influence and lying might look like, and how frequently they might occur.

Method

Sample

In collaboration with the Maricopa County Attorney’s Office, we obtained a sample of 398 victims represented across 252 cases of CSA occurring between January 2005 through December 2015 in Maricopa County. The County Attorney’s Office provided a list of all eligible cases. Cases were deemed eligible if they involved at least a single charge of: Sexual Conduct with a Minor (A. R.S.13–1405), Child Molestation (A. R.S. 13–1410), or Sex Abuse (A. R.S. 13–1404). We contacted and paid court reporters to share transcripts of cases; 73 court reporters were contacted and 47 responded (64% response rate). We received 214 complete victim’s testimonies across 142 cases (some cases included multiple victims); the remaining court reporters were non-responsive. Of these 214 testimonies, 134 were minors (age 17 years and under) (across 101 cases; M_{victim per case} = 1.33, SD_{victim per case} = .65), whereas the remaining transcripts involved young adults testifying about alleged victimization during their childhood.

For the present investigation, we examined the 64 testimonies of children aged 5 to 12 years old (M_age = 9.45, SD_age = 1.95, 11% males, 28% White, 30% Latinx, 14% other, and 28% unknown). Because children’s ability to understand subtle statements develops through adolescents, young children may be most susceptible to subtle attacks on credibility (compared to teenagers). Defendants (100% male) were the child’s parent or caregiver 30% of the time, another family member 30% of the time, a family friend or other familiar adult (e.g., coaches, babysitters, and neighbors) 32% of the time, and a stranger 8% of the time. Children alleged penetration or attempted intercourse in 25% of cases, oral copulation or genital contact in an additional 10% of cases, and less severe abuse in 52% of cases (fondling, exhibitionism, etc.). Ninety percent of cases resulted in a conviction of at least one charge.

For each transcript, we organized question-answer pairs into “lines of questioning.” We conceptualized this as a sequence of questions that surround a specific topic, or develop a line of argument (e.g., first line of questioning: “Q: You said the defendant touched you? A: Yes. Q: Where did he touch you? A: In my private parts.” Second line of questioning: “Q: Was the door open or closed when that happened? A: Closed.”). This unit of analysis was preferable because individual question-answer pairs may not sufficiently reveal an attorney’s intention to expose or contest suggestive influence.¹ Assessing the logical argument of a line of questioning helps clarify the intended meaning of subtle or ambiguous questions within the sequence.

Two coders independently read each transcript to identify lines of questioning. Discrepancies were discussed and resolved to reach 100% reliability. We identified 7,103 unique lines of questioning asked by prosecutors and defense attorneys across the 64 transcripts. Children were asked between 8 and 325 lines of questioning, with an average of 111 (SD = 66) lines of questioning per child. Prosecutors asked on average 72 (SD = 41) lines of questioning per child while defense attorneys asked on average 39 (SD = 33) lines per child.

Next, two coders independently read each line of questioning, assessing if the line asked about suggestive influence. We included lines of questioning about telling the truth (e.g., if the child was telling the truth in court or at disclosure), conversational influences (e.g., conversations that may have influenced the child’s disclosure, including coaching), incidental influences (e.g., experiences that may have influenced the child’s ability to describe abuse, such as witnessing abuse of others), and other, miscellaneous truthfulness/influence questions (e.g., motives to lie/disclose). After resolving discrepancies on which lines addressed suggestive influence, coders independently coded each line of questioning into subcategories based on emergent themes. They also independently coded lines of questioning as direct or indirect accusations. Direct accusations explicitly asked the child if they were telling the truth, if the abuse “really happened,” if someone told them what to say, or if incidental influences actually influenced their reports. Indirect accusations implied subtly that the child lied or was influenced by referencing potentially suggestive conversations or experiences, motives to lie, and other suggestive topics. After reaching K > .80 reliability on all categories and subcategories, the coders resolved 100% of discrepancies. Definitions and examples of categories are summarized in Table 1. Mean proportions of major categories of all lines of questioning are summarized in Table 2.²

Table 1.

Suggestive Influence Categories: Definitions and Examples.

Category	Definition	Example
Direct accusations
Truthfulness	Explicitly asks if the child told/is telling the truth at disclosure/interview/testimony; if the child lied/made up allegations/testimony; if the alleged offense really happened or never happened	Did you tell the truth about [X]?; When you talked to her, were you telling her things that really happened?
Conversational	Explicitly asks if someone told the child what to say, to lie, or to make up allegations	Did anybody ever tell you what to say about [defendant] touching your cookie?
Incidental	Explicitly asks if incidental influence caused disclosure/testimony	Are you saying [defendant] touched your butt because [other victim] said that?
Indirect accusations
Implied conversational influence
“Talk to/with”	If others talked to or with the child about abuse, testifying, or disclosing; if the child talked to many different people about abuse; if the child talked many times about abuse	Has your dad talked to you about your testimony today?
Practice	If others practiced, went over, or reviewed testimony/disclosure, what to expect in court, or documents/videos; child’s familiarity with the courtroom and procedures	Did you ever come down to court before to practice testifying?
Word choices	Where the child learned words/phrases used to describe abuse, including body parts, sex acts, euphemisms; sources of information about what happened	You said that the potty place was sort of the name your mommy gave to it, right?
Help remember/disclose	If someone helped the child remember or reminded the child what happened or their testimony; someone helped the child disclose/describe abuse, or figure out what to say	Did your mom help you remember certain things?
Implied incidental influence
Knowledge of allegations	Child’s knowledge of other victims’ allegations against the defendant; other victims disclosing to child; child overhearing disclosures of other victims	Did you know anything about him ever being accused or having any conviction for touching another child in the past?
Witnessed abuse	If the child saw the defendant abuse someone	Did you see [defendant] touch [other victim] in the wrong place?
Sexual education	Formal or informal education/knowledge about sex, good touch-bad touch, or stranger danger	Has your mom and dad ever talked to you about people not being allowed to touch your private areas?
Other subtle influence questions
Implied honesty problems	Truthfulness reminders; ability to differentiate the truth/what is real/lies/fantasy/make believe; child’s imagination; habitual lying	Did we talk about telling the truth?…Did we agree that that’s what we’re going to do today?
Motives to lie/disclose/testify	If someone forced or pressured the child to disclose/testify; if the child disclosed/testified to appease, help, or not disappoint someone; if the child disclosed/testified because of a motive to lie (without saying “lied” or “made up”)	And so there was opportunity for [your parents] to stay together by you telling them something happened to you; is that right?
Suggestive questioning styles	If an adult used suggestive questioning styles during a previous interview	Did you feel when you were talking to the lady police officer that there were right and wrong answers to her questions?
Miscellaneous	If the child practiced memorizing things, always does what others say, or lies habitually; if only remembers touching because watched interview video; if listened to other victim’s testimony; others	Did you have to, like in the reading tests, did you have to memorize things to say for today?

Table 2.

Proportions of Total Lines of Questioning Asked by Prosecutors, Defense Attorneys, and Overall.

	Children Asked at Least One Line of Questioning about Each Topic	Proportion of All Prosecutors’ Questions		Proportion of All Defense Attorneys’ Questions		Proportion of All Questions
	N (%)	N	M (SD)	N	M (SD)	N	M (SD)
Direct accusations	50 (78.1)	154	.04 (.05)	49	.08 (.07)	203	.03 (.05)
Indirect accusations	56 (87.5)	182	.04 (.04)	218	.10 (.12)	490	.05 (.05)
Implied conversational influence	45 (70.3)	76	.01 (.02)	111	.04 (.07)	187	.02 (.03)
Implied incidental influence	40 (62.5)	86	.02 (.03)	72	.02 (.02)	158	.02 (.03)
Other subtle influences	33 (51.6)	25	.01 (.01)	41	.01 (.02)	68	.01 (.01)
All credibility attacks	59 (92.2)	336	.07 (.07)	265	.13 (.15)	601	.90 (.07)
All lines of questioning	64 (100)	4,628	-	2,475	-	7,103	-

To assess differences in the frequency with which prosecutors and defense attorneys asked about different suggestive influence types, we conducted paired-sample t-tests for each category and subcategory. The mean frequencies of each type of suggestive influence line of questioning asked by attorneys and the t-tests results are presented in Table 3. Examples of lines of questioning highlighting differences across attorney types are presented in the appendix.

Table 3.

Mean Frequencies of Lines of Questioning Asked by Attorneys and Tests for Difference across Attorney Type.

	Prosecutor		Defense		Difference of Means Two-Tailed t-Tests Across Attorney Type (df = 63)				Overall
	N	M (SD)	N	M (SD)	M Difference (SE)	95% CI: Lower, Upper	t	p	N	Mean (SD)
Direct accusations	154	2.41 (2.88)	49	.78 (1.86)	1.63 (.39)	.84, 2.41	4.15	.00	203	3.17 (3.65)
Truthfulness	81	1.27 (1.58)	24	.38 (.98)	.89 (.19)	.52, 1.27	1.27	.00	105	1.64 (2.16)
Conversational	69	1.08 (2.06)	23	.36 (1.13)	.72 (.28)	.15, 1.28	2.54	.01	92	1.44 (2.44)
Incidental	10	.16 (.51)	2	.03 (.18)	.13 (.07)	−.01, .26	1.82	.07	12	.19 (.53)
Indirect accusations	182	2.84 (3.08)	218	3.41 (3.35)	−.56 (.44)	−1.45, .32	−1.27	.21	400	6.25 (5.37)
Implied conversational influence	76	1.19 (1.55)	111	1.63 (1.63)	−.44 (.24)	−.92, .04	−1.82	.07	187	2.81 (3.08)
“Talk to/with”	49	.77 (1.22)	67	1.05 (1.37)	−.28 (.20)	−.68, .12	−1.41	.16	116	1.81 (2.05)
Practice	14	.22 (.63)	16	.25 (.62)	−.03 (.09)	−.21, .15	−.35	.73	30	.47 (1.02)
Word choices	12	.19 (.47)	16	.25 (.67)	−.06 (.08)	−.23, .10	−.75	.45	28	.44 (.94)
Help remember/disclose	2	.03 (.18)	21	.33 (.91)	−.30 (.12)	−.53, −.06	−2.53	.01	23	.36 (.91)
Implied incidental influence	86	1.34 (2.03)	72	1.13 (1.76)	.22 (.27)	−.32, .76	.81	.42	158	2.47 (3.13)
Knowledge of allegations	39	.61 (1.15)	38	.59 (1.12)	.02 (.15)	−.28, .31	.11	.92	77	1.20 (1.94)
Witnessed abuse	37	.58 (1.31)	22	.34 (1.01)	.23 (.16)	−.08, .55	1.49	.14	59	.92 (1.97)
Sexual education	12	.19 (.50)	10	.16 (.48)	.03 (.08)	−.12, .19	.41	.69	22	.34 (.76)
Other subtle influences	25	.38 (.60)	41	.58 (1.07)	−.20 (.13)	−.46, .05	−1.61	.11	66	.95 (1.41)
Implied honesty problems	11	.17 (.38)	12	.19 (.43)	−.02 (.07)	−.15, .12	−.23	.82	23	.36 (.60)
Motives to lie/disclose/testify	8	.13 (.38)	14	.22 (.60)	−.09 (.08)	−.25, .06	−1.23	.22	22	.34 (.80)
Suggestive questioning styles	1	.02 (.13)	7	.11 (.54)	−.09 (.05)	−.20, .01	−1.76	.08	8	.13 (.65)
Miscellaneous	5	.08 (.32)	8	.13 (.45)	−.05 (.06)	−.18, .08	−.72	.47	13	.20 (.60)
Total sample	336	5.25 (5.09)	265	4.14 (3.82)	1.11 (.68)	−.24, 2.46	1.64	.11	601	9.39 (7.18)

Note. Significant effects in bold.

Results

We identified 601 lines of questioning (9% of all lines of questioning) about suggestive influence (i.e., truthfulness, conversational influences, incidental influences, and/or miscellaneous truthfulness/influence). Ninety-two percent (n = 59) of the children in our sample were asked at least one line of questioning about suggestive influence (M = 9.4 lines per child, SD = 7.2; range = 0–26). Thirty-four percent (n = 203) of suggestive influence lines of questioning were direct accusations, and attorneys asked 78% (n = 50) of the children at least one direct accusation. Indirect accusations were more common; 67% (n = 400) of suggestive influence lines of questioning were indirect accusations, and attorneys asked 88% (n = 56) of children at least one indirect accusation. These findings are consistent with prior literature (Denne et al., 2020; Stolzenberg & Lyon, 2014). Like Denne et al. (2020), we found that a substantial proportion of all attorneys’ questions asked about suggestive influence. Furthermore, we found that attorneys asked about suggestive influence in a greater proportion of cases (94%) than Stolzenberg and Lyon (2014) found in their study examining overt suggestive influence questions only. Our findings, therefore, confirm our expectation that attorneys asked about suggestive influence subtly. Below we describe how attorneys raised suggestive influence concerns, moving from direct accusations through the forms of indirect accusations.

Direct Accusations

We identified three subcategories of direct accusations: truthfulness, conversational influences, and incidental influences. Fifty-two percent (n = 105) of direct accusation lines of questioning asked explicitly if children were telling the truth (e.g., “Q: And did you tell the truth when you were talking to [forensic interviewer]?”). Additionally, 45% (n = 92) of direct accusation questioning asked explicitly about conversational influences, such as coaching (e.g., “Q: Did anybody ever tell you what to say about [defendant] touching your cookie?”). Finally, 6% (n = 12) of direct accusations asked if incidental experiences influenced children’s testimonies (e.g., “Q: Now, I know that you know what [victim 1] and [victim 2] saw because they told you right? A: Yes. Q: Okay. But the things that you told us that you saw today is that what you saw with your very own eyes?”). Prosecutors posed most of the direct accusation questioning (n = 154, 76%), including 77% (n = 81) of direct truthfulness, 75% (n = 69) of direct conversational influence, and 83% (n = 10) of direct incidental influence lines of questioning.

Indirect Accusations

Indirect accusations (n = 400 lines of questioning) were more common than direct accusations (n = 203), and they ranged in topic and subtlety. Forty-seven percent (n = 187) of indirect accusations subtly implied conversational influences, 40% (n = 158) implied incidental influences, and 16% (n = 62) made other attacks. Below, we describe these indirect accusations.

Implied conversational influence.

Implied conversational influence questioning subtly asked if someone coached the child or if conversations with others influenced their reports, decision to disclose, or description of abuse. Sixty-two percent (n = 116) of implied conversational influence questioning asked about the child “talking” to or with others about the case or testimony (e.g., “Q: …Now, before you said that the lady came to your house and talked to you guys, did you and your mommy talk about what was going on?”). Forty-two percent (n = 49) of “talk to/with” lines of questioning referenced children’s conversations with other alleged victims. Specifically, attorneys asked if children had “talked about” or “discussed” the abuse with others who had also alleged abuse against the defendant, typically siblings or friends (e.g., “Q: …Did you and [other victim] talk about what happened that night with [defendant]?”). Additionally, 26% (n = 30) of “talk to/with” lines of questioning asked about how often or with how many people the child talked about the case or testimony, topics some have suggested imply coaching (Lyon & Stolzenberg, 2015). For example, in a case involving an 11-year-old girl alleging a single instance of non-penetrative abuse by her father, the defense attorney asked the following line of questioning:

Q: Did you—you talked to [forensic interviewer] more than one time, right?

A: Twice.

Q: Twice. Did you talk to…the gentleman right here that’s sitting right there in the gray suit?

A: Yes, sir.

Q: [Name 1]?

A: Yes, sir.

Q: Okay. And have you talked with your foster mom…about different things?

A: Yes, sir.

Q: And also [Name 2]

…A: Yes, sir.

Q: You talked to her about some things?

A: Yes, sir.

Q: Okay. Did you—did you talk to [Name 3]?

A: Yes, sir.

Q: Before you came to court?

A: Yes, sir.

Q: How many times did you talk to [Name 3]?

A: Twice.

Q: You talked to her twice. How about the other lady with the glasses?

A: Yes, sir.

Q: You talked to her too?

A: Yes, sir.

Q: How many times did you talk to her?

A: Twice.

Q: And different people have been talking to you about just different things about the case, right?

A: Yes, sir.

Lines of questioning using “talk to/with” may not have been exclusively about coaching. Attorneys may have used this phrase to ask about disclosure (e.g., “Q: And I think you said that you remembered talking with your dad about what happened…Did I recall that correctly?”), the child’s emotional response (e.g., “Q: And do you talk about what happened or how you feel about it or something else?”), and trial preparation (e.g., “Q: Did you talk to anybody about … having to testify?”).³

Sixteen percent (n = 30) of implied conversational influence questioning asked about preparing for court (e.g., “Q: When you went over to the building next door to talk with these guys over here about you coming to court, you spoke with them about the kind of questions they would be asking you today, right? A: Yes. Q: They kind of told you what to expect and everything else?”). Fifteen percent (n = 28) of implied conversational influence questioning asked about children’s word choices, such as when and how the child learned particular words or phrases, or if others told the child what to say in court (e.g., “Q: And you had said that the potty place was sort of the name your mommy gave to it, right? A: Yes…Q: And do you now call it your potty place, too, like your mom does? A: No. Q: What do you call it? A: My private place. Q: Your private place. Okay. But as far as today and all the questions you’ve been asked, you’re explaining it the way mommy explained that area, right?”). Finally, 12% (n = 23) of implied conversational influence questioning asked if the child was helped to remember what happened, what to say, or to disclose (e.g., “…Q: And did it help to have people sometimes tell you what they thought really did happen? A: Yes. Q: Yeah? Did it help you remember things? A: Yes. Q: And sometimes did they tell you things that maybe you didn’t remember, but then after they told you it kind of made sense that maybe that is what happened?”). While prosecutors may ask practicing lines of questioning to discuss trial preparation, or ask about word choices to clarify terms, it is clear from the examples (all from defense attorneys) that lines of questioning about practicing, word choices, and others helping the child also subtly imply suggestive influence.

Implied incidental influence.

Implied incidental influence (n = 158, 40%) lines of questioning suggested that the child’s reports were influenced by incidental experiences, including knowledge of others’ allegations, witnessing abuse, and sexual education. Nearly half (n = 77) of implied incidental influence lines of questioning asked children if they knew about the allegations of other victims (e.g., “Q: You know that [other victim] made allegations against your grandfather, right? A: Yes. Q: Do you know what they are? A: Not really.”), or about other victims disclosing to the child (e.g., “Q: Okay. At some point did [other victim] come and tell you? A: No. Q: Did [other victim] come and tell you that [defendant] was touching her? A: No.”). Many (n = 26, 34%) of the lines of questioning about children’s knowledge of others’ allegations assessed when the child learned about allegations relative to their own disclosure (e.g., “Q: [Child], do you know if anyone else has—if your [defendant] has ever touched anyone else? A: Yes. Q: Who? A: My stepsister…. Q: Did you know that now or did someone tell you? A: My mom told me. Q: Your mom told you. Did you know that—did your mom tell you that after you told her what happened to you? A: She told me when she said I have to go to court.”). Rather than implying that someone told the child what to say, these lines of questioning implied that knowing that others alleged abuse, or what they alleged, could have influenced children’s reports.

Similarly, 37% (n = 59) of implied incidental influence questioning asked about children witnessing the abuse of other victims (e.g., “Q: All right, and at some point in time do you think that you saw [defendant] touch [other victim]? A: Yes.”). Additionally, 14% (n = 22) of implied incidental influence questioning asked about sexual education (e.g., “Q: Okay. Did they teach you that there’s certain places where people shouldn’t touch you? A: Yes. Q: Okay. What places are those that they taught you people shouldn’t touch you? A: My private and my chest.”). These questions revealed a concern that children who have witnessed sexual abuse or know about sex, such as from sexual education, could describe sexual encounters without being sexually abused.

Other subtle influence questions.

Finally, attorneys asked 66 lines of questioning that accused lying or influence in other subtle ways. Among these, 35% (n = 23) implied honesty problems, including reminding children to tell the truth or asking about others reminding them (e.g., “Q: And do you promise that today when you’re talking, we’ll only talk about things that really happened?”), questioning children to determine if they could differentiate truth from lies, or “what’s real” from “make believe” (e.g., “Q: Do you know the difference between telling the truth and lying?”);⁴ and asking about children’s vibrant imaginations (e.g., “Q: Do you believe you have a pretty good imagination? A: My dad thinks I do. But I’m losing it now because we’re always so busy….”). Thirty-three percent (n = 22) asked about children’s motives to disclose or testify, including if they felt pressured to testify (e.g., “Q: When you were being interviewed by [detective]…did you ever feel that he wanted something from you?”); if they disclosed or testified to help, appease, or not disappoint someone, such as a parent (e.g., “Q: Do you think that if you came into court, and said that [defendant] didn’t touch you, that that would disappoint your mom?”); and other motives to lie (e.g., “Q: Are you telling us today that [defendant] touched you because you’re mad at him for making you hold something up?”). Finally, eight lines of questioning asked about detectives’ and prosecutors’ suggestive questioning styles (e.g., “Q: Okay. Did you feel when you were talking to the lady police officer that there were right and wrong answers to her questions? A: Yes. Q: Okay. And is that because some of the questions she asked you gave you a hint as to what she thought the answer should be? [Objection by prosecutor].”).

Attorney Differences

Our analyses reveal variation across attorney type in the frequency and likelihood with which attorneys made direct and indirect accusations (Table 3). We also found that the quality of lines of questioning within categories differed by attorney. For example, while defense attorneys’ lines of questioning with the “talk to/with” phrase seemed motivated to subtly accuse coaching, prosecutors’ lines of questioning were slightly more overt (refer to Appendix). Similarly, prosecutors’ “practice” lines of questioning focused on children’s familiarity with the courtroom and courtroom procedures (e.g., “…Q: And have we talked about coming into court and talking? A: Yes. Q: In fact, did we actually go into another courtroom so you can practice sitting up there? A: Yes.”), and their “word choice” lines of questioning seemed intended to clarify terms rather than to subtly accuse suggestive influence (e.g., “Q: Okay…[W]hen you say [defendant] is the guy that raped you, what does raped you mean? A: Like he touched me in a spot that you don’t feel comfortable at. Q: Okay. And where did you learn the word “raped”? A: From my sister…”). Attorney differences in implied incidental influence lines of questioning were less obvious, though prosecutors and defense attorneys likely differed in their motives for posing these questions. For more examples comparing differences across attorney types, refer to Appendix.

Discussion

In this study, we examined how attorneys ask about suggestive influence to children testifying about alleged CSA. While researchers have examined attorneys’ explicit questions about suggestive influence (Stolzenberg & Lyon, 2014) and have found that asking about suggestive influence is common (Denne et al., 2020), ours is the first study to examine what subtle questions about suggestive influence look like and how frequently they occur. We hypothesized that questions about suggestive influence would occur often and that defense attorneys would ask more subtle questions than prosecutors. Our hypotheses were supported.

We found that attorneys commonly asked about suggestive influence. Nearly every child (92%) was asked at least one line of questioning about suggestive influence, and these represented 9% of all lines of questioning asked to children testifying. In addition, attorneys’ questioning about suggestive influence was mostly subtle. Overall, 67% subtly asked about suggestive influence. These findings are important. They show that prior research assessing overt questions about suggestive influence may have underestimated the frequency with which attorneys raise these issues and their relative importance in direct- and cross-examinations (Stolzenberg & Lyon, 2014). Concerns about suggestive influence, therefore, are evident in attorneys’ examinations of most children testifying about CSA, but attorneys more commonly address these concerns in subtle ways not previously examined by researchers.

Our analysis revealed questioning that ranged in subtlety. For example, questioning implied coaching by asking about the number and frequency of conversations between children and others, which highlighted opportunities to be coached. “Word choice” lines of questioning, by contrast, subtly suggested coaching by highlighting that the way children described abusive episodes was influenced; if others influenced the words children used, they also could have influenced the content of reports. Similarly, questioning about incidental influence raised credibility issues by suggesting that the influence on children’s testimonies may not have been direct or intentional. Questions about knowledge of others’ allegations, witnessing abuse, and sexual education revealed concern that children’s non-abuse experiences provide them sufficient knowledge to fabricate and describe sexual abuse even without others telling them what to say.

Within this realm of subtle questioning, we found that polysemous implicatures, including phrases like “talk to/with,” “practice,” and “help remember,” were common. Nearly 42% of all suggestive influence lines of questioning contained one of these polysemous implicatures. As noted above, polysemous implicatures are phrases whose implied meanings differ from their explicit meanings (Noveck, 2001). Because children’s ability to understand implied meaning increases with age (Sullivan et al., 1994; Wimmer & Perner, 1983), they may struggle to refute questions containing polysemous implicatures that subtly imply coaching or suggestive influence, even if there was no prior influence (Wylie et al., 2020).

These lines of questioning ranged significantly in what they “looked like.” For example, attorneys used the “talk to/with” phrase to ask about various topics, including conversations between children and other victims, how often and with how many people children spoke about the abuse or testifying, emotional trauma, and others. Similarly, attorneys’ questions about practicing or “going over” testimony addressed normal trial preparation as well as possible coaching or influence. Given the many possible meanings of these questions, it is likely that children misunderstood attorneys’ implied meanings and inadvertently acquiesced to subtle attacks on their credibility. Our findings, therefore, are important. They reveal that prior research focusing on overt attacks on credibility underestimated the frequency with which attorneys raise credibility concerns. Our systematic examination of indirect attacks revealed that attorneys attack children’s credibility more often than previously noted, and in many subtle ways.

Implications for Courtroom Practices

The frequency of subtle questioning about suggestive influence is problematic, as children testifying in court may not recognize the threat to their credibility. Specifically, theory of mind—the ability to recognize that others’ beliefs about the world can diverge from one’s own and to attribute false beliefs to others (Premack & Woodruff, 1978)—may be necessary to understand polysemous implicatures (Wylie et al., 2020). As theory of mind develops steadily into the teenage years (Sullivan et al., 1994; Wimmer & Perner, 1983), young children may lack the cognitive abilities needed to understand the polysemous implicatures in attorneys’ questions, which may result in children affirming questions suggesting coaching even when they were not coached. This is especially true when children perceive the person “helping” them positively, as they are likely to do with supportive adults, like parents or prosecutors (Wylie et al., 2020).

Our analysis revealed many instances where children affirmed they talked with others about their testimony, practiced, or used words others told them to use. Children may affirm polysemous implicatures implying coaching when they are factually true. A child may say “yes” in response to the question “Did you talk with your mom about testifying?” because they did, but not in the way implied. A “yes” response would be a true response, but the ambiguity of the question requires more explanation to clarify that the child was not coached. A “yes” response may misrepresent the child’s conversation with their mother. Furthermore, our analyses revealed that attorneys’ motives for asking questions with polysemous implicatures varied; many of their questions may not have been about suggestive influence at all. So, to provide accurate accounts of events, children must understand the multiple meanings of attorneys’ questions and correctly interpret their motives. To the extent that children lack the cognitive abilities to do this—particularly maltreated children whose cognitive development is often delayed (Font & Berger, 2015)—polysemous implicatures can lead to ambiguous and contradictory testimony, which complicates assessment of credibility, especially when these phrases are common.

Prosecutors seemed attentive to these concerns. Prosecutors were more likely (both in frequency and proportion) to overtly ask about suggestive influence, posing 75% of direct accusations. In contrast, defense attorneys frequently asked questions that may have capitalized on children’s developmental limitations; 82% of all defense attorneys’ questions about suggestive influence were subtle. They were especially likely to ask about implied conversational influences, posing nearly 60% of these questions, including 58% of “talk to/with” lines of questioning and 91% of “help remember/disclose” lines of questioning, which subtly implied children’s reports were coached or influenced by conversations with others. Defense attorneys also asked about implied incidental influence, such as children’s knowledge of others’ allegations, to imply that children’s reports were influenced by this knowledge. Questions about witnessing the abuse of others and sexual education likewise may have implied that these experiences enabled children to fabricate reports. Prosecutors asked subtle lines of questioning about these topics too, though their motives likely differed. For example, prosecutors may have asked “talk to/with” lines of questioning to “remove the sting” (Ohler v. United States, 2000; Stolzenberg, 2012) of defense attorneys’ questions about coaching in cross-examination or ameliorate coaching concerns after defense attorneys raised them.

The frequency of defense attorneys’ subtle attacks is not surprising. Unlike some jurisdictions where attorneys must explicitly state their position that the child is lying or influenced during cross-examination (Hanna et al., 2012), attorneys in the United States are instructed to merely imply that the child’s testimony is false or the product of influence (Myers, 1986, 1994). Defense attorneys working in the United States, therefore, may be motivated to subtly accuse children of influence. Subtle attacks, however, are problematic, because children who misunderstand the possible meanings of indirect accusations (e.g., questions using “talk to/with” phrases), or attorneys’ motives, may inadvertently admit influence or lying, even when their testimonies are truthful. So, allowing for subtle methods that children are developmentally ill-equipped to refute may falsely undermine children’s credibility.

Children’s—and jurors’—interpretive tasks are further complicated by prosecutors who often use the same subtle phrases to establish, rather than attack, children’s credibility. Furthermore, as prosecutors do not systematically ameliorate suggestive influence concerns overtly, particularly concerns about incidental influences, the adversarial process may not sufficiently counteract the effects of defense attorneys’ subtle attacks. Subtle questioning about suggestive influence, therefore, creates unnecessary challenges. Some foreign jurisdictions have addressed these challenges by prohibiting subtle questioning altogether (Hanna et al., 2012). Nevertheless, legal traditions in the United States may limit the ability of jurisdictions to implement similar policies. Subtle, developmentally insensitive questioning of children is likely to remain common in CSA trials conducted in the United States. As such, the onus is on prosecutors to limit their own use of subtle questioning, to recognize the subtle questions of defense attorneys, and to mitigate any credibility concerns that may arise by asking clear, straight-forward questions that children of all ages can understand.

Limitations, Future Directions, and Conclusions

Our study highlights the problematic nature of courtroom questioning practices for children. Using age as a proxy for cognitive development, we revealed how questioning that is insensitive to witnesses’ cognitive and linguistic abilities may produce confusing, inconsistent, or misleading testimony. Age, however, is not the only factor associated with witnesses’ ability to communicate effectively in court. Researchers find that lower socio-economic status and experiences of abuse are associated with delayed development (Font & Berger, 2015; Hackman & Farah, 2009); race and gender likewise influence communication styles and developmental trajectories (Coates, 2015; Farkas & Beron, 2004). Unfortunately, due to the racial and gender homogeneity of our sample, as well as a lack of information on victims’ socio-economic status and history of (non-sexual) abuse, we were unable to examine the role of these factors in attorneys’ questioning about suggestive influence. Researchers should explore how these characteristics and experiences influence questioning practices and their effectiveness.

While using transcripts allowed us to identify the form of questions attorneys use in the field, we could not assess how children interpreted attorneys’ subtle attacks on their credibility. Assessing how children responded to suggestive influence questioning was also outside the scope of the current investigation. Researchers can continue to assess how children interpret and respond to subtle attacks on credibility in the laboratory, as well as how subtle questioning about suggestive influence affects children’s accuracy and credibility. Furthermore, while we can sometimes infer attorneys’ motives, we cannot know for sure what they intended. Interviewing attorneys about their strategies for addressing suggestive influence concerns, or their awareness of subtle strategies to attack credibility, could help us understand attorneys’ meanings.

Relatedly, while we expect that questioning about suggestive influence affects children’s credibility, we were unable to determine the direct effects of such questioning on jury’s credibility assessments or verdicts. How questioning children about suggestive influence affects case outcomes is unknown, and an important avenue for future research. Researchers could also examine how potential jurors understand subtle questioning. Given the multiple meanings and motives for asking the questioning we identified, jurors may not always interpret these questions as asking about suggestive influence. Assessing how jurors interpret subtle questioning would contribute to our understanding of the effectiveness of subtle attacks on credibility.

Overall, our study has contributed to our understanding of how attorneys ask children about suggestive influence. As expected, attorneys’ accusations of lying, coaching, and influence were usually subtle. Attorneys subtly question children about suggestive influence in ways not previously examined, including where children learned the words they used to testify and incidental influences. Importantly, prosecutors and defense attorneys were equally likely to ask subtle questions about these topics. Both attorneys, therefore, commonly asked children about suggestive influence in ways that were developmentally inappropriate, easy to misunderstand, and difficult to refute, potentially exacerbating misunderstandings and misinterpretations. Furthermore, while prosecutors asked most children explicitly if they were coached or were telling the truth, they rarely asked about incidental influence explicitly. So concerns about incidental influences may remain even after prosecutors remedy coaching and truthfulness concerns through overt questioning. Attorneys concerned with establishing the veracity of children’s reports should, therefore, address coaching, truthfulness, and other influence concerns with overt questions, as subtle questioning can exacerbate rather than ameliorate credibility concerns when children allege sexual abuse.

Author Biographies

Suzanne St. George, MS, is PhD student in the School of Criminology and Criminal Justice at Arizona State University. Her research focuses on sexual violence, including theories of perpetration, rape myth acceptance, consent communication, and sexual assault case processing. She is also interested in how race, gender, and sexuality intersections influence perceptions of and responses to rape.

Colleen Sullivan, MS, is a PhD student in the Law and Behavioral Sciences Program at Arizona State University. Her research focuses on CSA and the methods used to interview children, family law, and extralegal factors that influence legal decision-making.

Breanne E. Wylie, MA, is completing her PhD in the Lifespan Development program at Brock University. She is interested in methods used to question children in the court system. Specifically, she is interested in developing recommendations for questions that should be avoided with young children and which questions children are best able to demonstrate their knowledge in response to.

Kelly McWilliams, PhD, is an assistant professor in the Department of Psychology at John Jay College of Criminal Justice. As a developmental psychologist her work focuses on children’s eyewitness memory and forensic interviewing. Through laboratory and field research, she has examined current investigative and prosecutorial practices in the legal system, and their influence on the behavior and responses of child witnesses.

Angela D. Evans, PhD, is a developmental psychologist and associate professor at Brock University. She studies the influence of children’s social and cognitive development on their moral understanding and behavior. She uses experimental methods to systematically study children’s understanding of deception and their actual deceptive behaviors both in the laboratory and in the field.

Stacia N. Stolzenberg, PhD, is an assistant professor in the School of Criminology and Criminal Justice at Arizona State University and a core faculty for ASU’s program on Law and the Behavioral Sciences. Trained as an applied developmental psychologist, Stolzenberg takes an interdisciplinary approach to study children’s reports of maltreatment and then applying her work to the criminal justice system.