Abstract
Confronted with a difficult, unexpected, or confrontational question, an expert witness may answer by attempting to overwhelm the questioner with words, sometimes highly evasive ones, that avoid, rather than actually address, the question asked. Such a discursive response is sometimes called a “waffle,” as in “The expert's answer was a waffle.” This review notes some examples of this phenomenon and attempts to categorize them in a meaningful way. An ancillary goal of this discussion may be to aid experts in focusing their answers.
[That was] authentic frontier gibberish!
So says a character in Mel Brooks' parody of a Western movie, Blazing Saddles, after another character has spouted a stream of incomprehensible verbiage from his toothless mouth through his beard.
The instructions commonly given to expert witnesses on how to answer a question under oath at deposition and trial include pausing to replay the question in the mind, thinking it through, and making a responsive answer. In deposition, the goal is to preserve a clear record for later use at trial. The expert also attempts to avoid painting him- or herself into a corner and avoids creating bases for the use of the testimony for possible impeachment at trial. Trial testimony carries the added requirement that a lay jury be able to understand clearly, and, theoretically, be persuaded by, the testimony.1,2
Under some circumstances, this ideal is not met. An expert may tumble into the pitfall of, as the proverb states, “operating the mouth before the brain is in gear.” This functional difficulty may produce the wandering, prolix, discursive, and ultimately evasive answer known as a “waffle”—a result with several different causes. The subject has not been widely discussed in the literature, though some teaching videotapes3 address it.
Origins of Waffling
Waffling may result from the expert's failures to listen to the questions actively and thoughtfully, to think through the answer in advance, or to grasp with sufficient clarity and coherence the point of the question. Another wellspring of the waffle is the expert's wish for appearance's sake to be seen as readily answering the question, even though the answer has not been well planned out. Physicians often chafe at the restriction of “just answer yes or no,” and may rebelliously wax prolix.
The anxiety and stress of the deposition or trial setting alone, of course, may lead to the expert's rapidly blurting out ill-considered answers. Yet another example is when the expert, dismayed at having only a vague sense of the answer, attempts to throw more and more words at the question in the hope that one, or perhaps the sheer volume of the totality, will strike a responsive chord. A similar waffle may occur when the expert inserts numerous qualifiers into the answer because he or she has not prepared sufficiently to have a clear idea of his or her own opinion. A related process is when the expert tries to remake the question being posed into the question he or she wishes had been asked, instead of answering the original.
A more problematic reason for the waffle may be a conscious attempt to avoid answering the question for fear of weakening the case for the retaining side (i.e., to preserve a partisan, rather than objective, opinion). Occasional experts believe that their role is to oppose, or refuse to concede, any question on cross-examination. It is as though “good testifying means never saying yes to the other side” (Meyer D, personal communication, July 17, 2006). Similarly, some experts act as though they'd rather die than say “I don't know” on the witness stand.
And, in a worst-case scenario, a venal witness or “hired gun” may waffle in an attempt to provide a vague basis for a fundamentally untenable opinion. In such situations, a venal expert may have a prepared “stump speech,” rather than a responsive opinion, that they give repeatedly (e.g., “It is clear malpractice, and that is my professional opinion.”).
Note that these different factors may be quite difficult to tease apart, since, under the stress of cross-examination, an honest witness may find his or her thoughts thrown into confusion, leading to an apparently thought-disordered response that is not an attempt to obfuscate, but an effort to make some response to a stress-inducing question. In addition, novice experts, proofreading their own early deposition transcripts, describe the experience as akin to hearing one's recorded voice for the first time: strange, alien, and hard to recognize as one's own.
Some responses are distinguishable from the actual waffle. Lengthy, complex, detailed, tentative, and even ambiguous answers are not necessarily waffles if they respond to the question rather than evade it.4 Certain cross-examination questions, indeed, may require an extended narrative response that actually addresses the question, no matter how inartfully the latter is asked.
The Current Survey
In this survey, some classic forms of actual waffling are presented for analysis. As a courtesy, the actual sources are not identified. The author attests in good faith that these are direct quotes from either deposition or trial testimony—and thus matters of public record—without modification. A few comments of context will be provided to make the issue at hand comprehensible. All identifying data have been eliminated to preclude recognition. In addition, to preserve anonymity and dignity, the excerpts are not in the order of the above listing of types.
Example 1
In a murder case that resulted in a liability claim against the treaters, an expert was asked to comment on the significance of the psychological tests that were performed on the perpetrator as aids to the treaters' decision-making: The idea of, condensing his assessment, overlooking other matters such as the degree of his persecutory thinking and some of the other specific kinds of notations about his thinking, condensing it into an appraisal, summarily, of low to moderate, doesn't have any foundation in the behavioral sciences. And it was again, my professional opinion notwithstanding what appears to be a quite a valid administration of these tests, thoughtful administration of these tests, thoughtful communication of the results, a very interesting useful information that certainly Dr. X [the treater] would have benefited from, that the idea of consolidating it into low to moderate risk is where, based on what he detailed in the facts of his findings, there is no basis for connecting one with the other.
In this example, the response ostensibly starts to address, and then backs away from, so many dimensions of the question that the response is actually difficult, perhaps even impossible, to follow. The core point appears to be that a level of risk assessment could not be, or should not have been, derived from those tests. Naturally, other interpretations are possible.
Example 2
In a case involving predicting the dangerousness of a mental patient, where liability is claimed for the patient's violent act several years after an assessment, the expert is asked about validity of the long-range prediction of dangerousness: And to explain. If I learned what I learned about what was being said at the [scene of the violence] in [date] about what was being communicated in [date] in [later date] or at other times, then the richness of what would have been available to [treating physician] would have been far more detailed, far more vivid, and in certain instances, perhaps, would have enabled him to appreciate the urgency immediately, as well as to guide treatment planning at the time. There was real detail available and out there that a psychiatrist, if I knew that, if I was treating and I heard the things that I heard as a treating psychiatrist, I mean, would absolutely have influenced what was going on in my treatment relationship with the patient.
This example appears to be suggesting that some missing data may have had an impact on the treater's decision-making, but other interpretations of the quotation are possible. The expert seems to be tending toward stating one idea—the possible utility of some information—but is expressing it poorly by drawing it out repetitively with stops and starts. Moreover, using one's own practice as the example does not convey the standard of care.
Example 3
In this civil case of emotional injury one of the plaintiff's complaints is that he becomes fatigued faster. The deposing attorney asks: Q: Can you date the onset of [the plaintiff's being] fatigued faster?
The responsive answer to this question might call for either a date or a context. The expert replies: A: According to him, he felt that there was, that related to ebb and flow a little bit, but really minor—he himself didn't look at it as—he just looked at it as stress and pressure, having to do with when [other plaintiff] was more—well, certainly when she was depressed and suicidal and had the onset of her major depression. And then when they were, in particular, in the heat of dealing with [a therapist] and their discussions about him changing his job, he pretty close to minimized there being much of a problem, maybe just a little concentration problem, maybe just a little distraction, but pretty coherent clear and capable at work until after [date]. And then—do you want to move to that?
Q: When after [date] did this problem of becoming fatigued faster occur?
Note that such repetition of the question often signals that, from the attorney's viewpoint, the question has not been answered; of course, sometimes such repetition is an attorney's ploy to invalidate the answer, but here, no actual answer has been given: A: Well, he felt both—he sees as his role protector, manager, the guy who doesn't make a lot of demands, who is there to control and all. And given his major sense of self and his role as a husband and father, and all, he saw this as a major threat. So—and it got worse as—I would say another major bump was, as they tried to see about becoming active and getting a way to deal with this situation, and kept getting rebuffed, that really preoccupied him. But—and then finally, when they got representation and then the next increment would be going public with them. Because in their minds there were the two tracks. They were worried that that might occasion much more risk. He had those thoughts, the dreams, and the thoughts about [a movie allegedly similar to case] and stuff like that. So, he knew, he knew, knew, there would be potential risk in confronting this situation. He wasn't really thinking so much about the trial and what the legal process was going to be. I don't think he really—I don't know how advised he was about that and how prepared he was. To some degree maybe, but it was worrying that [defendant] was going to get mad at him and do something and to theirs.
The deposing attorney made one more attempt to get a specific answer and then gave up. This example is difficult to parse, but it clearly does not provide anywhere near an answer to the question. One gets the impression that attempts are being made to give a comprehensive (albeit incoherent) clinical picture of the patient, rather than responsively to identify the time at issue. Here, an expert with relatively little forensic experience apparently was not tuned into the narrow and specific forensic Q and A that should characterize depositions.
Example 4
A defense expert in a suicide malpractice case (where the treater might have hospitalized a patient who committed suicide as an outpatient) has given evasive and even obstructive answers throughout the deposition, making it necessary for some questions to be repeated multiple times. The following dialogue occurs: Q: After reading [treater's] depositions and reviewing chart notes, are you telling me that you cannot give an opinion as to what [patient's] suicide risk was on [date]?
Again, the answer might be a yes or no, with or without an explanation, and, if possible, an opinion. The expert replies: A: I think, knowing [treater's] relationship with the patient, the information that he has on working with her over the years and her family, and the information that is here and the testimony that he gave in his deposition, he felt—and the information would validate—that her suicide risk was low at that time or else he would have done more. I think there was some concern of which he was under the impression I think from the notes that she was staying with family and he'd had conversations with her family about that.
Note the hidden circular reasoning: if he knew X, he would have done more; he did not do more, ergo, he did not know X. This is, of course, the core negligence issue as to whether he got the needed data. Q: I am wondering if you're able to give an opinion, looking at the chart notes and everything you have reviewed, as to what you think [patient's] suicide risk on [same date as above].
A: I don't think it's appropriate for me to give an opinion that is—I can give you an opinion on just the information that I have in front of me, and I do not think that applies at all to the opinion that [treater] may have had because of his relationship with the patient, and the family play a great role in what that decision would be.
Note that this last answer begins with the idea that he cannot give an opinion about suicide risk, shifts to the opinion that he can give about an assumed difference between the treater's opinion and the expert's, makes the causal connection for that difference to the relationship between treater and patient and family, and ends with the treater's decision-making which, though quite relevant to the question of negligence in the case as a whole, provides no answer to actual question: suicide risk on a particular date. Overall the response has a defensive feel to it.
These answers, among other problems, have shifting frames of reference in the middle. This trait guarantees confusion and a muddy record.
Example 5
The same case as in Example 4: Q: Doctor, isn't it true that, nationally as well as locally, the options in treating a suicidal patient include hospitalization, supervision, medication and counseling?
A: In a general sense those are not always options available. There are some times when hospitalization is not available. Sometimes when there is no therapist, other people available, or there's times when there's no family available. And so those factors would influence that.
This kind of broad and explicitly inclusive question represents what might be called the “no-brainer” or throwaway question. The answer is “yes”; nothing is lost by saying so. Of course, any given case may pose exceptions, but to respond as the expert does sounds argumentative and quarrelsome. This waffle appears to result from the fear that some problematic concession is being sought, and the expert evades it. As a general principle, throwaway questions should be thrown away; argumentative or discursive answers only weaken the expert's credibility. If asked whether patients sometimes commit suicide despite one's best efforts, no other answer is as honest, clear, and substantive as “Yes.”
Example 6
Same case. Q: [Treater] testified that on [date] there was no indication for hospitalization. Do you agree with him or disagree?
A: I would state that I think hospitalization would be something that should have been considered as an option among many options.
Q: [Treater] testified that there was no indication for hospitalization. Would you agree or disagree?
A: If there are other options available that are appropriate, then there would be no indication. And he felt there were other options that were appropriate. So there was no indication to hospitalize her at that time, as how I would interpret that statement.
The last response takes some rereading, especially given the curious first sentence, but it is actually close to responsive, though still structurally a waffle. The witness is avoiding giving his own opinion as asked; instead, he attempts to interpret the treater's reasoning.
Example 7
The issue is the restraining of a patient. Q: Does the standard of care require restraint of this patient?
A: When you have a patient who is this out of control and depressed, it's a very serious situation, and you have to respond and give the patient what he needs or otherwise you'll have a really bad situation.
Q: But does the standard of care require that this patient be restrained?
A: I've already answered that.
It is unclear whether the witness really believes he has answered the question, despite its being repeated twice. It seems more like evasive waffling, but other interpretations are possible. In fact, in most of the examples, it remains unclear whether the experts themselves felt that they had, indeed, answered the questions.
Example 8
This question of patient committability was put to a plaintiff's expert in a case of liability for the patient's violence. Q: So you think after his discharge from [hospital X] in [month] of [year], he should have been [that is, the standard of care required that he be] involuntarily admitted to a hospital before [month] of [year]?
A: It is my opinion that one of the possibilities that should have been seriously considered would have been an involuntary hospitalization. It certainly should have been considered. I am not in a position to tell you that that would be the only choice [Ref. 1, pp 54–5].
In this response, the expert seems to tend toward a “yes” answer, but then backs away from it and actually avoids answering. Note that even short answers can be waffles.
Example 9
The following waffle is one continuous run-on sentence. In this example, a patient has escaped from the hospital and committed suicide. The plaintiff's expert answers the query as to the bases for his opinion that treatment was below the standard of care. The entire answer took four full deposition pages, but this excerpt is representative. [Expert:] The standard of care in my professional opinion was breached in that, once the patient left, the mental state and what's gone on in that patient's mind is very uncertain, that this is patient with some history of, a reasonable history, actually, of unpredictability; he gets frightened, he has taken in despair 10 lithiums some years back, took some blood pressure pills one time in [city], goes all the way to [another city], we don't know whether he stops or doesn't stop and get [drug] or not, but specifically there is a lot of despair and a great deal of thought disorganization in the patient, and where I believe the standard of care was breached was that the patient, an emergency petition ideally would have been, reasonably would have been, rather than ideally, reasonably should have been issued so that the patient could have been brought back for reassessment in terms of their thinking and what possessed the patient to leave, an hour before that or less signs a 3-day statement [a formal request to leave the hospital from a voluntary admission] and then just disappears [Ref. 1, p 55].
The deponent twice begins with the phrase, “the standard of care was breached … ”, an apparent beginning of a responsive answer, but both times turns in other directions. The answer also shifts among time frames, symptoms, and possible treatment responses. This scattershot response is a classic waffle, apparently aimed at flooding the answer with words in the hope of being seen as actually responsive.
Example 10
The last example is in response to the question of deviations from the standard of care in a discharge in a Tarasoff-type case. It's a consequential piece of behavior that creates the most essential elements of a treatment plan, which makes sure the patient is safe within a structured environment, and that includes they would be safe within or without the community, because the treatment will eventually take place if it can at all within a less restrictive alternative that is community based, but that doesn't mean that it's without supervision [Ref. 1, p 56].
In this waffle, the deponent seems to vacillate between positions (e.g., within or without, community-based but not without supervision) trying to work both sides of the street, as it were. Of course, this avoids committing to a position.
Discussion and Recommendations
At the outset, I reviewed some of the dynamic issues that might lead to waffling in both its benign and malignant forms. No matter what its origin, waffling essentially fails to achieve the clarity necessary to reach a lay jury.
The jury itself may be a useful barometer of one's own chance of waffling. Most instructions to experts from their own attorneys or various other sources in the literature emphasize the importance of facing the jury when speaking to them. The members of the jury will often telegraph to the testifier via body language the effect he or she is having, and the waffle is no exception. The witness who notices that the jurors are frowning, glancing at each other, shaking their heads a bit or manifesting the dreaded “glazed-eye sign” should consider stopping in midpresentation and saying something like: “Let me try to say this more clearly.” If the situation is as suspected, the jurors should now show facial and body language expressive of relief.
An attorney once provided me with a helpful critique by suggesting: “Treat your sentences as though they were paragraphs.” This idea implied that delivering each sentence as a separate entity, rather than stringing them along, would make it easier for jurors to take them in. A brief pause at the end of a statement also allows an opportunity for the content to “sink in.”
In addition, some very basic problems can typically be solved by, in fact, returning to the basics:
Listen to the question, focus on the question, be sure you understand it, think through the answer in a rehearsing manner, then give it in an organized fashion.
Break up long thoughts into short phrases.
Finish your thoughts.
Finish your sentences.
Do not shift frames of reference in midanswer; if a different perspective is subsequently called for, finish the first answer and then begin another.
Do not be afraid to state that you cannot answer a particular question.
Do not be afraid to ask for rephrasing of the question for clarity.
Do not be afraid to state that you do not understand a question.
Do not be afraid to say “I don't know.”
In addition to these fundamentals, anticipatory preparation of expected cross-examination answers is extremely worthwhile. These principles are likely to be useful in general; however, the waffling response may not be altered or improved by these basics, since it may not spring from incompetence or inexperience but from the intention to evade and avoid. Though commonly a technique of the venal expert or hired gun, other explanations and influences are possible as indicated herein, including overidentification with plaintiff, defendant, or respective attorney. Expert witnesses are well advised to avoid waffling and look with suspicion at waffling by opposing experts.
Acknowledgments
The author thanks members of the Program in Psychiatry and the Law, especially Archie Brodsky, BA, Donald Meyer, MD, and Harold Bursztajn, MD, for critical comments, and Ms. Ellen Lewy for assistance with the manuscript.
- American Academy of Psychiatry and the Law