dissenting:
The majority’s decision to reverse appellant’s conviction for attempted second-degree cruelty to children1 marks the first time in over five decades that the doctrine of “inherent incredibility” has been used to set aside a criminal conviction in the District of Columbia.2 Indeed, this is the first time that this court has applied the doctrine since we became the final authority over District law after court reorganization in the early 1970s. I believe that the doctrine ill fits this record, and therefore I dissent, respectfully.
I. Facts
In this very sad case, appellant’s wife, Louisetta Koh, told appellant after a heated argument that she would be taking away her daughter and their, toddler son (D.S.) to North Carolina the next day. Appellant’s sister, Donna Robinson, heard the ruckus and saw the couple tussling inside the front door of the' Robinson house (where Ms. Koh and the children were living). Ms. Robinson told appellant to leave. He refused, grabbed the high *544chair with D.S. strapped into it, and a melee ensued. Appellant and Ms. Koh ended up on a couch, with Ms. Koh on her back on the bottom, appellant facing her on top, and D.S. in the middle, still in the high chair, facing his father. Ms. Koh kept demanding that appellant let go of D.S. while Ms. Robinson was trying unsuccessfully to pull him off. Ms. Robinson’s husband heard screaming, called 911, then came downstairs and began to' hit appellant. Even the downstairs tenant came upstairs and joined the fray.
Apparently Ms. Koh wriggled off the couch, leaving appellant — who would not let go — still hunched over the couch, enveloping D.S. in the high chair. According to Ms. Robinson, appellant had a “tight” grip on D.S. with his “weight on the baby.” When two police officers arrived in response to the 911 call, one confirmed that appellant was “laying on top of the baby,” who “was faced up and crying.” The other officer testified that appellant was holding D.S. “very tightly” while the child was “crying hysterically.” The officers “more than twice” ordered appellant to release D.S. and eventually had to use “reasonable” force “to .., get him to comply.” The trial court found that appellant had held D.S. in a “dangerous and reckless” position “for an extended period of time”— more specifically, “for many minutes even when the police arrived!,] continuing to do so[ ] despite being asked to let the baby go.”
II. Standard of Review
The standard for reviewing a challenge to sufficiency of the evidence following a bench trial is well established: “[W]e consider all the evidence in the light most favorable to the government, according deference to the fact-finder to weigh the evidence, determine the credibility of the witnesses, and draw all justifiable inferences of fact.”3 We will not reverse a conviction unless “the government presented no evidence upon which a reasonable mind could find guilt beyond a reasonable doubt,”4 and we will not discard a trial court’s findings of fact unless they are “plainly wrong” or “without evidence to support [them].”5 However, “if the testimony of a witness is inherently incredible under the circumstances,” we will not defer to the trial court’s credibility determination.6
As the majority acknowledges,7 the test we use to gauge whether testimony is “inherently incredible” is “very stringent.”8 Thus, we may find testimony “inherently incredible” only when it
[1] can be “disproved ... as a matter of logic by the uncontradicted facts or by scientific evidence,” or [2] when the “person whose testimony is under scrutiny made allegations which seem highly questionable in the light of common experience and knowledge, or [3] behaved in a manner strongly at variance with the way in which we normally expect a similarly situated person to behave;”[9]
*545For the reasons that follow, I do not believe that the majority has met this “very' stringent” standard. •
III. The Majority’s Analysis
In its analysis, the majority focuses on three interlocking trial court findings: (1) that appellant “held the. baby, who was strapped in a highchair, in a ‘tight grip’ (2) that he “put his ‘weight on the baby* (or on ‘some of the baby5), with the child ‘pinned between his body and the high chair on top of a couch’ and (3) that he did so “for many minutes, still *very tightly’ gripping the baby.”10 Inherent in addressing these findings is resolution of two aspects of the trial court’s analysis which the majority finds problematic.- The first is the court’s heavy reliance on allegedly inconsistent testimony by, Ms. Robinson when describing the tightness of appellant’s grip, and the severity of his weight, on D.S. The second is the court’s failure to recognize physical limitations, discernible from appellant’s knees on the floor, his torso leaning on the couch, and the child’s location in a high chair — all of which allegedly mitigated the gravity of the risks from appellant’s grip and weight. Based on these supposed shortcomings, the majority concludes that Ms. Robinson’s credited testimony was “inherently incredible,” and therefore requires rejection of the trial court’s findings predicated on that testimony.11
At the outset, I believe- it is important to emphasize what my colleagues do not deny, namely, that there can be a grave risk of bodily injury to a toddler who is strapped into a high chair, picked up and gripped tightly by his father, then flipped over backward onto a couch on top of the child’s mother, with the father’s weight on the child-in-chair for an extended period of time while the mother wriggles away and family members try — then the police use force — to make the father let go. The question here, therefore, is limited to whether there was credible evidence that appellant’s grip was tight enough, his weight heavy enough, and his resistance long enough to have created that grave risk.
A. “Tight Grip”
As to Ms. Robinson’s testimony about appellant’s grip on D.S., only one of the case- law criteria for an “inherently credible” finding is relevant, number [2]: “allegations which seem highly questionable in the light of common experience and knowledge.”12 My colleagues find inherently incredible Ms. Robinson’s testimony that appellant had a “tight ... grip on the baby” because- she also testified that she had been standing behind appellant, where she could see only his back and the back of his head, adding that “you really couldn’t see the baby[;] ... [y]ou 'really couldn’t see the baby good, unless you knew the baby was there, you wouldn’t have known that he had that tight of a grip on the baby” (emphasis added). Ms. Robinson’s “unless” clause, however, makes all the difference here. Ms. Robinson witnessed the entire sequence of events, including what happened before appellant fell on top of D.S., hiding him from view. She’ saw appellant grab D.S. while D.S. was strapped into the high chair and pull the child-in-ehair close to his chest. She knew that D.S. was still there, strapped in as before, when appellant fell over with him onto the couch. Thus, her' testimony about the tightness of appellant’s grip after D.S. went out of view was informed by her *546observation of that grip before he went out of view, reflecting the downward pressure from the face-forward fall. Unlike my colleagues, therefore, I am not convinced that Ms. Robinson’s testimony that appellant had held D.S. in a “tight” grip contained internal inconsistencies that rendered that testimony inherently incredible. Indeed, Ms. Robinson’s “tight grip” testimony was all the more credible because it was corroborated by Officer Rodd’s testimony that appellant “was holding the child very tightly.” It was further corroborated by testimony elaborating on the family’s unsuccessful efforts to physically force appellant to release his grip, as well as Officer Keenan’s testimony that a “reasonable amount of force” was necessary to make appellant let go.
Furthermore, the court’s “tight grip” finding is not undermined, as the majority would have it, by Ms. Robinson’s testimony that appellant had “a full highchair ... in his grasp,” not just D.S. alone. The high chair itself (no one has argued that a tray was on it) may have mitigated some of the risk of injury to D.S. — a fact that Ms. Robinson herself admitted on direct examination. (“The only thing that I could see that might have even helped a little bit was the highchair.”) The trial court, however, not only was aware of that testimony but also was confronted by defense counsel’s argument that the high chair’s “frame” protected D.S. Nonetheless, the court concluded, in light of all the testimony, that appellant’s grip around the child-in-chair was tight enough, when added to other risk factors, to put D.S. at grave risk of bodily injury. My colleagues and I were not present to hear the testimony; only the trial court was in a position to judge the witnesses’ demeanor (including intensity), as well as their words. Under these circumstances, therefore, I see no room for us to say that the court’s eyes and ears were defective in failing to find the “tight grip” testimony inherently incredible. To the contrary, given our standard of review, I cannot conclude that the court’s “tight grip” finding was “plainly wrong or without evidence to support it.”13
B. “Weight on the Baby”
The majority’s other reliance on “inherent incredibility” in the trial court’s fact-finding negates the risk from appellant’s “weight” on D.S., namely, the court’s failure to recognize the mitigating effects of the high chair, coupled with appellant’s weight-bearing knees on the floor and torso leaned against the couch. Straightaway I note that appellant never raised the knees-and-torso argument in the trial court, a curious omission that my colleagues downplay in accepting the argument on appeal. Presumably trial counsel saw nothing there.
In any event, the trial court’s failure to discern sua sponte the significance of posture over testimony would fall only within “inherent incredibility” criterion number [1]: disproof of the finding “as a matter of logic by the uncontradicted facts or by scientific evidence.”14 No scientific evidence is involved here, and, for me, it is much too far a stretch to say that Ms. Robinson’s testimony regarding appellant’s weight, on which the trial court relied, was rendered inherently incredible simply “as a matter of logic” by testimony regarding the high chair and appellant’s posture. The high chair and posture testimony, on which the majority relies, did not spawn “uncontested facts” that undermined the “tight grip” and “weight” testimony with ironclad, night-follows-day logic.
*547In emphasizing police officer testimony that appellant had his “knees on the ground” with his torso “leaning on the couch,” my colleagues conclude that. a “substantial portion” of appellant’s .weight “necessarily” was borne by his “knees (and/or the-couch),” not entirely'by-D.S.15 However, in my judgment, this knees and couch surmise — tantamount to appellate court fact-finding — does not effectively .negate the trial court’s record-based finding that enough of appellant’s weight was “on the baby” or on “some of the baby” to cause risk. That an unspecified, though arguably substantial, portion of appellant’s weight was borne by his knees, and perhaps the couch, does not necessarily mean that an insignificant portion of his weight was on D.S. as appellant leaned with upper-body weight over the boy and chair, gripping them tightly.
What does the majority have in .mind by “substantial” weight held off of D.S.? 30% ? 51% or more? What percentage of appellant’s weight on D.S. was too little to help create grave risk? . We are not told.
Because Ms. Robinson saw the entire chain of events that led to appellant’s falling on top of D.S.,:she was in a position to assess' whether appellant was placing his weight on D.S. She acknowledged, as noted above, that the high chair may have partially protected the child, but she emphasized nonetheless that appellant’s weight (meaning from the scenario his upper body weight) was on D.S. to an appreciable extent, based on her observation of all'; that had happened. I see no reason why a reasonable fact-finder could not find this testimony credible.16 The fact that neither police officer expressly testified that appellant had placed his' “weight” on D.S. does not nullify Ms. Robinson’s testimony. To the contrary, although the officers lacked her overall perspective, arriving well after appellant had fallen on the child-in-chair, their testimony tended to corroborate, • not undermine, Ms. Robinson’s observations. Recall the police testimony that appellant was “laying on top of the baby,” evidence tending to support the trial court’s finding that D.S. was bearing at least some of appellant’s upper-body weight. ■
Contrary to my colleagues’ apparent premise, the trial court need not have found that all or even most of appellant’s weight was on D.S. to find that appellant’s weight was a factor — among others — contributing to a grave risk of bodily injury. The obvious size differential between appellant and his sixteen-month-old son lends credence to that finding.
Finally, my colleagues offer no analogous case law supporting reversal. In each case they rely on, ante at 539 & n. 9,17 reversal for inherent incredibility was *548premised on obvious evidentiary inconsistencies and not, as here, on mere interpretation of coherent testimony (contradicted only by appellant himself and his wife).
C. “For Many Minutes’.’
The legal significance of the trial court’s findings as to appellant’s “tight grip” and “weight” was intensified by the court’s further finding, uncontested by the majority, that appellant persisted with pinning the child down “for an extended period of time” — “for many minutes.” .In fact, according to the court-, appellant held on to the little boy. “even when the police arrived” and “despite being asked [by the officers] to let the baby go.” Eventually, according to police testimony, the officers had to use.- a “reasonable amount of force.... [I]t took a little while to get him off.”
The, trial court found that the risk to D.S. from this police action, necessitated by appellant’s unyielding hold, contributed to the risk of bodily injury. Said the court; “This is extremely dangerous ignoring [the] police order to release the baby ... and continuing to hold the baby-Both officers testified they had to physically get him ... off of the baby.” The need for the police' to use force to extricate D.S. not only indicates how tight appellant’s grip must have been but also how the additional time,- with additional tussling over the child, intensified the risk. Because appellant resisted for “many minutes” the force directed at pulling him- off of D.S., the court could have permissibly inferred, without speculation, that, appellant naturally would have been trying to tighten his hold while reinforcing his weight on the child-in-chair, as the child continued to cry hysterically — as much an indication of pain (another signal foretelling risk of bodily injury) as of fright.18
D. -Trial Court Ruling
We have said that the statute proscribes a “grave risk of bodily injury, not a risk of grave bodily injury”; thus, the factfinder must focus on the mere “likelihood of injury,” not on the degree of potential injury.19 Consistent with that distinction] this court has intimated that “grave,” in modifying risk, means “substantial,”20 but' not necessarily a risk likely to produce significant harm. Although neither the trial court nor the majority discusses relevant definitions, I assume that both the court and my colleagues have presupposed that, for conviction, there must be evidence of a “grave” or “substantial” risk- of a bodily injury at least more than de minimis. Personally, I do not understand why a grave risk does not imply a risk of grave injury;21 but, even if a risk, to be “grave,” *549must portend a grave injury, not something less, I would not agree that the trial court’s findings of fact essential to appellant’s conviction are plainly wrong or without evidence to support them.
I turn, therefore, to the nature of the risk itself In addressing the trial court at the end of the hearing, defense counsel raised perhaps thé most likely risk and questioned whether the child could “possibly suffocate” with a' high chair “frame around himself.’' The court replied, “I’m not making a specific finding whether the child was going to suffocate.'... There could have been many ways the child could have been harmed under these circumstances because the defendant refused to release the baby. So I’m not here to make a physics [finding] as to whether the chair was providing sufficient protection in -order to prevent suffocation . ”
It is unfortunate that the trial court did not indicate that, yes, suffocation was among the possible risks, for surely the so-called “physics” .of the situation — tight grip and pressure from upper body weight for many minutes — suggested that obvious possibility. But the trial court’s response did not appear intended to discount that risk, a risk of bodily injury that assuredly was evident (in the words of the case law)22 “in the light of common experience and knowledge.” As I understand the court’s findings and judgment, therefore, the court foresaw the possibility of a variety of injurious (I’d say grave) outcomes and apparently perceived those risks to be obvious — and reasonably so. (I venture that the court would have had in. mind not only suffocation but also, for example, facial abrasions, broken bones, and other crush-related internal injuries.) The fact that the risks apparently did not result in actual injury to D.S. is a fortunate outcome that in no way undermines the trial court’s verdict and -judgment under a statute criminalizing-only “risk.”23
IV. ‘ Conclusion
I believe that the trial court, in crediting testimony from Ms. Robinson and Officers Keenan arid Rodd (while discounting the testimony pf appellant, Mr. Robinson, and Ms. Koh) did not engage in “mere speculation,” as my colleagues put it,24 to convict appellant of attempted second-degree child ‘cruelty.. The majority’s,two examples of “inherent incredibility” — Ms. Robinson’s testimony regarding ' appellant’s “tight grip” and “weight on baby” — fail to satisfy the statutory and case law criteria 'essential to withholding deference to the trial court’s findings in a bench trial. That testimony is not inherently incredible based on alleged inconsistency or on al*550leged mitigation of the risk by the high chair or by the couch (a posture argument not raised at trial). On this record, therefore, I cannot join my colleagues in holding that the evidence was insufficient to support appellant’s conviction for attempted second-degree child cruelty.
iji 'Jfi #
I must add that, although I cannot join ■with my colleagues, I am sympathetic with the outcome, if only because it is not clear to me why the government brought this case. We appear to have here a fracturing family, with a father apparently devoted to his young son and a mother apparently remorseful that appellant was on trial. In the end, moreover, although actual injury is not an element of the crime, it is clear that, despite the evident risk from appellant’s behavior, the child was not harmed. All that said, I have no idea whether government forbearance, rather than prosecution, would have helped heal this family’s relationships, and I have no basis, let alone prerogative, for second-guessing the prosecutor’s decision-making. Moreover, I am well aware that the statute is justifiably intended to protect children by proscribing behavior that causes grave risk, before eventual harm. I write this final paragraph, therefore, with some hesitation, intending only to disclose that judges often wish they knew more of what has been going on in a case than the record discloses, especially when family distress is before us.
. D.C.Code § 22-1101(b)(1) (2012 Repl); see ante at 543 n. 1.
. See Jackson v. United States, 353 F.2d 862, 867-69 (D.C.Cir.1965) (trial court’s determination of probable cause to arrest reversed because of fact-finding predicated on "inherently incredible” police testimony).
. Jones v. United States, 67 A.3d 547, 549 (D.C.2013) (internal quotation marks omitted).
. Russell v. United States, 65 A.3d 1172, 1176 (D.C.2013) (internal quotation marks omitted).
. D.C.Code § 17-305(a) (2012 Repl.).
. Robinson v. United States, 928 A.2d 717, 727 (D.C.2007).
. Ante at 543.
. In re 491 A.2d 490, 496 n. 8 (D.C.1985).
. Payne v. United States, 516 A.2d 484, 494 (D.C.1986) (quoting In re A.H.B., 491 A.2d at 496 n. 8 (quoting Jackson, supra note 2, 353 F.2d at 867)).
. Ante at 545.
. Ante at 539-43.
. See text accompanying supra note 9.
. D.C.Code § 17-305(a).
. See text accompanying supra note 9.
. Ante at 541.
. It is true, as my colleagues point out, ante at 535, that the trial court noted that Ms. Robinson had violated the rule on witnesses by discussing her testimony with her husband before he testified. As a sanction, the court drew an inference that both Robinsons were biased against appellant. ‘ At' the same time, however, as the majority also notes, ante at 536, the trial court — even in drawing the adverse inference — found Ms. Robinson to be "a very compelling and credible witness,” a perception that this court has no convincing basis for discounting.
.Jackson, 353 F.2d at 864, 867-68 (police officers’ testimony that they had received tip with detailed description of appellant, allegedly possessing heroin, ruled inherently in- • credible because testimony was "internally] contradict[ory]” and "contrary to human experience,” and officers initially had arrested and searched the wrong man); Farrar v. United States, 275 F.2d 868, 869-70 (D.C.Cir.1959) (testimony of complaining witness that appellant had threatened her with knife and "constantly pressed it against her” ruled inherently incredible because of complainant’s repeated acknowledgment that she had been *548“looking at him” but never saw knife, and further because neither complainant nor her clothing was. "marked,by a knife”); Davis v. Commonwealth, No. 1637-99-2, 2000 WL 1182013 at *3, 2000 Va.App. LEXIS 612 at *5-*9 (Va.Ct.App. Aug. 22, 2000) (series of conflicting statements and testimony by. complaining witness so “internally self-contradictory” that her testimony alleging sexual assault was inherently incredible).
.See Newman v. United States, 49 A.3d 321, 325 (D.C.2012).("This is not a case where the trier of fact crossed the bounds of permissible inference and entered the forbidden territory of conjecture or speculation.”) (internal quotation marks omitted).
. Lee v. United States, 831 A.2d 378, 382 (D.C.2003).
. Coffin v. United States, 917 A.2d 1089, 1092 (D.C.2007) (adverting to Iowa statute); see Dorsey v. United States, 902 A.2d 107, 113 (D.C.2006) (referring without' discussion to "grave or substantial risk”).
,. In the context most relevant here, "grave” means ‘.‘likely to produce great harm or danger,” as in "a grave mistake.” Merriam-Web*549ster’s Collegiate Dictionary 1174 (10th ed.1993) (the edition closest in time preceding enactment of the child cruelty statute); see The American Heritage Dictionarx of the English Language (3d ed.1992) ("grave” means "[draught with danger or harm,” as in "a grave wound”). On two occasions when interpreting the cruelty statute we have relied on a dictionary definition. See Mitchell v. United States, 64 A.3d 154, 156 (D.C.2013) (citing Alfaro v. United States, 859 A.2d 149, 157 (D.C.2004) (quoting definition of "cruel” in Webster’s Seventh Collegiate Dictionary 200 (1966))).
. See supra note 9 and accompanying text.
. Appellant also challenges the sufficiency of the evidence supporting the trial court’s finding that he acted with the state of mind necessary to support conviction under the second-degree child cruelty statute, i.e., "intentionally, knowingly, or recklessly.” D.C.Code § 22-1101(b)(1) (2012 Repl), Given the credited, and in my view credible, testimony (which the majority does not challenge) concerning the duration of appellant’s actions that placed D.S, at grave risk of bodily harm, I also would conclude that the trial court’s finding that appellant acted recklessly was supported by sufficient record evidence. -
. Ante at 539.