United States v. Christopher SammonsUnited States v. Christopher Sammons
Appeal from the United States District Court for the Southern District of Ohio at Columbus. No. 2:19-cr-00107-1—Sarah Daggett Morrison, District Judge.
COUNSEL
ARGUED: Kevin M. Schad, FEDERAL PUBLIC DEFENDER‘S OFFICE, Cincinnati, Ohio, for Appellant. Kevin Koller, UNITED STATES ATTORNEY‘S OFFICE, Cincinnati, Ohio, for Appellee. ON BRIEF: Kevin M. Schad, FEDERAL PUBLIC DEFENDER‘S OFFICE, Cincinnati, Ohio, for Appellant. Kevin Koller, UNITED STATES ATTORNEY‘S OFFICE, Cincinnati, Ohio, for Appellee.
SUTTON, Chief Judge. In online chats with two undercover officers, Christopher Sammons shared child pornography, requested videos of his correspondents
I.
On March 11, 2019, undercover FBI Agent Aaron Hurst made an online post feigning interest in incestuous sexual abuse. An anonymous user, later identified as Sammons, responded. He explained that he babysat his six-year-old niece every two weeks and offered to share videos of future interactions with her. Sammons also sent Agent Hurst child pornography. In return, Sammons repeatedly requested videos of Agent Hurst abusing his (fictitious) daughter.
Agent Hurst continued the correspondence as he tried to uncover the anonymous user‘s identity. On Saturday, March 23, Sammons stated that he would see his niece that coming Thursday and described plans to record himself abusing her. To show Agent Hurst that his niece was “real,” Sammons shared three photographs of her wearing a bathing suit. R.159 at 52. He also told Agent Hurst that he had confirmed his niece was off school on Friday, leaving “more time with her alone.” Id. at 54.
On Monday, March 25, Agent Hurst identified Sammons as the anonymous user and forwarded the case file to the Columbus FBI. Detective Brett Peachey of the Columbus Child Exploitation Task Force took action. Posing as a mother with two young daughters, Officer Peachey contacted Sammons that afternoon. Sammons shared a photo of his niece and again detailed the abuse he inflicted on hеr. Sammons also promised that, if Officer Peachey stayed up late on Thursday, he would send a video of him abusing his niece.
That Thursday, Officer Peachey and an FBI team detained Sammons when he returned home from work. They read Sammons his Miranda rights, and he agreed to be interviewed. Sammons initially denied wrongdoing but began to acknowledge his conduct as Officer Peachey confronted him with the gathered evidence. Sammons admitted to corresponding with the personas adopted by Agent Hurst and Officer Peachey and provided his chat username and phone password. He also confessed to exchanging child pornography online, including with Agent Hurst. While he denied penetrating his niece, he admitted to рulling her underwear to the side to take a picture of her genitals and to sharing it online. Sammons added that he had photographed her naked in the bathtub, and that “she enjoyed it” and “wanted [him] to do it.” R.163 at 221. The interview lasted approximately 45 minutes. It ended with Sammons’ arrest.
Due to the messaging application‘s data-retention policies, officers recovered only 30 days’ worth of Sammons’ messages. The records showed that Sammons sent and received child pornography and distributed photos of his niece wearing a bathing suit. But the records did not include images or videos of Sammons abusing her.
A grand jury charged Sammons with one count of child sexual exploitation,
Sammons’ trial defense focused on the argument that a compliant personality and PTSD—stemming from childhood abuse and a tour in Afghanistan—led him to falsely confеss. Sammons took the stand and admitted to distributing child pornography. But he denied abusing his niece and argued that he had falsely confessed because he was “willing to say what [Officer Peachey] wants me to say to just get it over with.” R.163 at 166. The court instructed the jury that “people, for various reasons, may admit to crimes they did not in fact commit,” and that the jury should evaluate the confession based on all the evidence presented at trial. R.160 at 84. The jury convicted Sammons on each count, and the court imposed a 300-month sentence.
II.
Sammons challenges his convictions on three grounds: that one-on-one messages do not amount to notices inviting child pornography; that the court erroneously excluded Dr. Bresler‘s testimony; and that insufficient evidence supported his child-sexual-exploitation conviction.
A.
Meaning of statute. Federal law forbids “any person” from “mak[ing], print[ing], or publish[ing] . . . any notice or advertisement seeking or offering” child pornography or child sexual exploitation.
In passing the lаw, Congress did not define “notice,” requiring us to look to its ordinary meaning at the time. When Congress enacted
In оther places in the United States Code, Congress uses “notice” in a similar way. It deploys the term to cover purely private notices directed to one person in non-public settings. See, e.g.,
In everyday speech, a “notice” frequently refers to a private one-on-one communication. That is particularly so when the subject matter of the communication is confidential or potentially embarrassing. Commonplace examples abound: layoff notices, lease termination notices, past due payment notices, notices of rent increase, and notices of violation.
The reality that some, perhaps many, notices are publicly disseminated to more than one person does not alter the term‘s meaning. General language often encompasses commonplace and atypical examples alike, making it important not to “narrow a provision‘s reach by inserting words Congress chose to omit“—here, by inserting “public” before “notice.” Lomax v. Ortiz-Marquez, 140 S. Ct. 1721, 1724–25 (2020).
Another aspect of the statute shows that “notice” covers one-on-one communications. It does not refer to a “notice or advertisement“; it refers to “any” such “notice or advertisement.”
The subject of the notice—child pornography or child sexual exploitation—reenforces this conclusion. Only the most brazen criminal would write notices “seeking or offering” participation in criminal activity in public and broadly distributed ways, as opposed to doing so in covert and surreptitious ways. A statute criminalizing only notices for child sexual exploitation flown behind blimps, published in The Wall Street Journal, or otherwise publicly shared would not prove particularly effective. Surely Congress sought to deter notices for child pornography likely to work, not to deter those unlikely to receivе a response.
That the notices may be “mailed” also supports this understanding.
The broader statutory scheme bolsters this interpretation. At the time of
Not only does a public dissemination requirement lack support in the language and history of the law, but it also would create exceedingly difficult problems of application over when a message has a sufficiently large audience to count as a public notice. Would a text message sent to two people count? What about an identical letter sent to three? Or an internet post accеssible by only five? The fog and blur hovering over a supposed line between public notices and private notices cautions against such an interpretation. See RJR Nabisco, Inc. v. Eur. Cmty., 579 U.S. 325, 342 (2016). That approach also would shield the most dangerous abusers who prey upon the most vulnerable victims. Why would Congress exempt the hardest-to-detect abusers engaged in one-on-one communications while punishing those less careful and more indiscriminate in their solicitations?
A near consensus of our sister circuits runs in the same direction. The Ninth Circuit, in interpreting
The Eleventh Circuit, it is true, has taken a different approach. See United States v. Caniff, 955 F.3d 1183 (11th Cir. 2020) (per curiam). It acknowledged that the definitions of “notice” do not have a public dissemination requirement. See id. at 1188–89. Even so, it reasoned that the term was ambiguous because dictionary usage examples, such as “put[ting] a notice on a door,” and “insert[ing] a [notice] in the newspaper,” referenced public notices. Id. No one doubts, however, that a notice may be public; the debate is whether it must be and above all whether “any notice” must be. That is a heavier lift, one not facilitated by some examples of public notices. Just as we would not conclude that only animals can be “vehicles” simply because “mule[s]” are listed as an example, see Vehicle, Webster‘s Third New Int‘l Dictionary 2538 (1993), we should be careful in assuming that some usage examples occupy a field.
The Eleventh Circuit also thought that context supported its position. Caniff, 955 F.3d at 1189-91. But the most salient pairing in the sentence—that Congress referred to “any notice or advertisement“—undermines the idea that notice refers only to public announcements or invitations. Advertisements, after all, frequently will be public, and Congress‘s use of “or” suggests
The court was also troubled by the phrase “make[] . . . any notice,” reasoning that an ordinary speaker would more readily use the expression “give notice” when referencing a private notice. Id. at 1189–90. Maybe; maybe not. But the salient question is how the ordinary author of federal laws, not any speaker, would use language in writing a broad statute designed to cover all manner of solicitations for child pornography and necessarily using a lot of words in the process. Making notice, like making law, may not be the most pleasing formation; offering or giving notice might work better, just as passing or enacting a law might roll off the tongue more smoothly. But these possibilities do not give us license to ignore
It is not even clear that “make” has a significant role to play in the statute. The Sentencing Guidelines contain provisions that the courts and litigants alike interpret as covering the same ground as
Consistent with this view, the word “make” when used in connection with a noun like “notice,” simply is another way, if a wordy way, of using the verb form of the noun. In other words, to “make notice” means to “notify.” That is just what the Court said, and did, in Janus Capital Group, Inc. v. First Derivative Traders, 564 U.S. 135, 142 (2011) (“The phrase at issue in Rule 10b-5, ‘[t]o make any . . . statement,’ is thus the approximate equivalent of ‘to state.‘“). So too here.
The Eleventh Circuit ultimately concluded that the statute was ambiguous and applied the rule of lenity. Caniff, 955 F.3d at 1191–92. But lenity comes into play only when a statute remains “grievous[ly] ambigu[ous]” after exhausting all of the traditional tools of statutory interpretation. United States v. Castleman, 572 U.S. 157, 173 (2014); see also Shular v. United States, 140 S. Ct. 779, 789 (2020) (Kavanaugh, J., concurring) (explaining that lenity applies only where a “court can make no more than a guess as to what the statute means“). Grievous ambiguity requires a plausible interpretation favorable to the defendant in near equipoise with the interpretation advanced by the government. See Johnson v. United States, 529 U.S. 694, 713 n.13 (2000). The Eleventh Circuit, which acknowledged engaging in a “back-and-forth, tennis-match-ish analysis” focused on potential flaws with the government‘s interpretation, Caniff, 955 F.3d at 1191, did not make such a showing, and Sammons cannot do so either.
It is not enough simply to show awkwardness to demonstrate a cognizable ambiguity. Ambiguity-triggering canons, like the rule of lenity, require a workable competing definition of a term or phrase. See TMW Enters., Inc. v. Fed. Ins. Co., 619 F.3d 574, 580 (6th Cir. 2010); United States v. Bryant, 663 F. App‘x 420, 422 (6th Cir. 2016). But that is doubtful here. A competing definition of “make” invoked by the Eleventh Circuit, though not Sammons, that it means to create or “cause to exist.” But that leads to a strange interpretation of the statute. It would mean that “make notice” covers only the mere creation of notices. Why would Congress punish the creation of notices about seeking or offering child pornography without any requirement that they go to someone? Sure, “make notice” may sound awkward by itself. But it is less awkward in connection with the setting in which we find it, a broad set of listed prohibitions used in legislative drafting, not everyday speech, that is simply another way of saying “notifying.”
Sammons offers other ways to counter our conclusion. He insists that “notice or advertisement” is a term of art that requires public dissemination when used as a single phrase, regardless of what “notice” or “advertisement” might mean independently. Consider several statutes, Sammons says, that use similar (or identical) language. See, e.g.,
Noting that words are known by the company they keep, Sammons claims that “advertisements” are necessarily public and that “notices” should be too. His argument depends on two premises—the first questionable, the second unsupportable. While “advertisement” often will require a public audience, that is not always true. Webster‘s Dictionary, for example, offers several definitions contemporaneous with the statute‘s passage, only one of which refers to a public audience. See Advertisement, Webster‘s Third New Int‘l Dictionary 31 (1993) (“the action of advertising: a calling attention to or making known,” “warning, admonition,” “an informing or notifying,” and “a calling to public attention“).
But even if that were not the case, this canon casts little light Sammons’ way. The
Another canon—the canon against surplusage—undercuts Sammons’ approach. The Supreme Court has “noted time and time again,” that we are “obliged to give effect, if possible, to every word Congress used.” Nat‘l Ass‘n of Mfrs. v. Dep‘t of Def., 138 S. Ct. 617, 632 (2018). But interpreting “notice or advertisement” to apply only to public communications essentially reads “notice” out of the statute. If advertisements must be public, as Sammons insists, and “notice” is cabined to public communications, it is difficult to determine what work “notice” does in the provision. Congress, it is true, may sometimes err on the side of redundancy and “employ[] a belt and suspenders approach” to ensure its aims are met, Atl. Richfield Co. v. Christian, 140 S. Ct. 1335, 1350 n.5 (2020), but it rarely employs multiple words (joined by “or” and рreceded by “any“) to adopt a narrow approach.
Sammons claims that the verbs “make[], print[], or publish[],”
Sammons adds that the notice offense originated in a section labeled “Advertising Offenses Related to Sexual Exploitation of Children.” Pub. L. No. 99-628, 100 Stat. 3510 (1986). From this beginning, Sammons infers that Congress intended the statute to punish only public inquiries, not all notices. This argument relies on the same questionable proposition that all advertisements necessarily are public. More problematically, the argument does not work on its own terms. If we accept for the sake of argument that section headings and statutory titles can shed light on the text of a law, that light reveаls a red signal, not a green one. The statute‘s title says that the offense encompasses “solicitations.” See id. (“An Act To amend title 18 of the United States Code to ban the production and use of advertisements for child pornography or solicitations for child pornography, and for other purposes.“). The title‘s reference to “solicitations” or “advertisements” parallels the statute‘s reference to “notices” or “advertisements.”
Legislative history, Sammons submits, shows that Congress never intended the statute to apply to “private internet communications” because “the internet did not exist” when the statute was enacted. Appellant‘s Br. 17. But unless otherwise limited, laws apply “to technology unknown when the operative words took effect.” Scalia & Garner, supra, at 86–87. That is why, for example, an old statute prohibiting theft covers the theft of an iPhone. See Pa. Dep‘t of Corr. v. Yesky, 524 U.S. 206, 212 (1998) (A statute‘s application “in situations not expressly anticipated by Congress does not demonstrate ambiguity. It demonstrates breadth.“). By the same token, new technology may escape old statutory language. That is why, for instance, a federal excise tax Congress intended to apply to every long-distance phone call perished in the face of changed industry billing practiсes. OfficeMax, Inc. v. United States, 428 F.3d 583 (6th Cir. 2005) (explaining that telecoms no longer charged based both on the duration of the call and the distance between the parties, as required for the tax to apply). Either way, a statute‘s application to advancing technology is ultimately a question of “the provisions of [the] law[],” not the “concerns of [the] legislators” who enacted it. Id. at 593.
B.
Exclusion of expert testimony. District courts serve a “gatekeeping role” in screening expert testimony to ensure that only reliable evidence goes to the jury. Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 597 (1993). Rule 702 of the Federal Rules of Evidence requires district courts in the first instance to determine whether the expert‘s testimony “will help the trier of fact,” is “based on sufficient facts or data,” is produced from “reliable principles and methods,” and “reliably applie[s] the principles and methods to the facts of the case.” In assessing reliability, courts look to a methodology‘s testability, exposure to peer review, error rate, and general acceptance in the relevant scientific community. Id. at 593–94. District courts have “considerable leeway” in performing their gatekeeping duties, requiring us to review their admissibility decisions for abuse of discretion. Conwood Co. v. U.S. Tobacco Co., 290 F.3d 768, 781, 792 (6th Cir. 2002).
The court did not abuse its discretion in concluding that Dr. Bresler‘s proposed testimony was not the product of reliable methods, reliably applied—the core of the Rule 702 inquiry. See United States v. Gissantaner, 990 F.3d 457, 463 (6th Cir. 2021). Dr. Bresler concluded that Sammons possessed a compliаnt personality and therefore may have falsely confessed. But, on a closer analysis, Dr. Bresler‘s methodology amounted to little more than vouching for Sammons’ credibility with the weight of a Ph.D. See United States v. Adams, 271 F.3d 1236, 1245–46 (10th Cir. 2001).
A look at the primary method Dr. Bresler employed, the “Gudjonsson Compliance Scale,” reveals its flaws in the criminal context. R.97 at 20. The technique consisted of reading Sammons twenty true-or-false questions about his tendency for compliance—such as, “I try to do what is expected of me” and “I tend to give in to people who insist they are right.” Id. at 60. Sammons knew the point of the not-under-oath questions—bolstering his false-confession defense—and thus had ample motive to prevaricate.
Dr. Bresler‘s effоrts to “validate” Sammons’ answers do not overcome this problem.
The validity index did not validate Sammons’ answers anyway. It instead showed “significant elevation,” indicating that Sammons may well have provided false answers. R.74 at 10. Dr. Bresler initially rationalized Sammons’ elevated score as a product of the stress of incarceration and “psychological maladjustment.” Id. When pushed on the elevated score at the Daubert hearing, however, he shifted. There, he attacked the very validity index he had selected, claiming it suffered from overinclusion and poorly measured deception. He instead proposed other validity indices—not mentioned or calculated in his report—that putatively supported Sammons’ truthfulness. Later in the hearing, Dr. Bresler reversed course again, this time stating that when assessing the possibility of feigning, he “would certainly start with the [first validity indеx].” R.97 at 54. The district court was well within its discretion to conclude that this confusing and contradictory testimony “undermine[d] any credibility” that the validity index could lend to the compliance questions, ultimately leaving them “with none.” R.101 at 15.
Nor did Dr. Bresler‘s second attempt at validation—asking Sammons’ wife to answer the same 20 questions about him—fix the problem. She had a credibility deficit of her own, as his supportive wife and as someone who would testify as a defense witness at the trial. See id. at 15 n.4 (recognizing a “similar[]” capacity to “manipulat[e] the test results“). Even if taken at face value, moreover, her answers did not validate Sammons’ answers. Sammons answered “yes” to all but one question, while his wife answered yes to only 11 of 20 (and could not provide answers for 6).
Dr. Bresler‘s lack of credibility compounded the methodology‘s lack of reliability. In Dr. Bresler‘s initial report, he found Sammons both “very compliant” and “highly suggestible.” R.74 at 11. The “suggestibility” conclusion was based on yet more questions, which, like the compliance questions, do not attempt to detect deception. When confronted at the Daubert hearing, however, Dr. Bresler abandoned suggestibility entirely. He claimed that Sammons had only a “mildly” elevated suggestibility score, disclaimed any reliance on the suggestibility questions, and stated that Sammons had not falsely confessed “because of suggestibility.” R.97 at 57, 93. To smooth over this mid-case pivot, Dr. Bresler pushed the envelope of candor in saying that he had never believed that suggestibility played a role in Sammons’ confession. Compare id. at 93 (“I didn‘t want to cherry-pick what I put into the report. That‘s why I described the [suggestibility questions].“), with R.74 at 11 (“[T]he [suggestibility questions] indicate that Mr. Sammons is a highly suggestible individual.“).
Stripped of jargon, Dr. Bresler‘s method boiled down to finding 20 different ways of asking a defendant—who understood that acquittal hinged on undermining his confession—whether he was compliant. To “validate” those answers, Dr. Bresler asked yet more questions, which may well have indicated dishonesty. When that method of validation proved less than
If we shift from the general reliability problems with this testimony to the more specific Daubert factors for assessing reliability, a similar conclusion follows. Take testability. Dr. Bresler conceded that the compliance questions could not produce a probability that a subject falsely confessed. In the absence of any such empirical scale, there is no way to measure falsifiability. See Daubert, 509 U.S. at 593 (explaining thаt “falsifiability, or refutability” define the scientific method). Nor, due to the obvious risk of feigning, does the record indicate that scientists have tested, or even can test, the theory that defendants who answer “yes” to more of the compliance questions falsely confess more often. Turn to peer review. The government‘s psychologist could not recall a single study addressing feigning on the compliance questions. Nor could Dr. Bresler identify any study addressing his proposed method of validating Sammons’ answers with the personality questions. What about error rate? Dr. Bresler acknowledged that, because the compliance questions could not produce quantifiable conclusions in the first place, they have no known error rate. Nor can general acceptance save them. Some psychologists may be willing to use them. But even Dr. Gudjonsson cautioned against their use in criminal cases due to their “vulnerability to feigning.” R.157 at 24.
Sammons insists that the testimony of the government‘s psychologist renders this a battle of the experts for the jury to resolve. Not necessarily. The government‘s psychologist never examined Sammons and testified not to propound his own diagnosis but to explain the flaws in Dr. Bresler‘s methodology. When an astronomer points out the flaws in the methods of an astrologer, or a chemist those of an alchemist, a jury need not resolve the dispute. See Kalamazoo River Study Grp. v. Rockwell Int‘l Corp., 171 F.3d 1065, 1072 (6th Cir. 1999) (explaining that court did not “choose sides” in a “battle of the experts” by excluding one expert‘s testimony as unreliable).
Even if the court properly excluded Dr. Bresler‘s testimony, Sammons adds, the exclusion violated his constitutional right to present a complete defense. See United States v. Reichert, 747 F.3d 445, 453 (6th Cir. 2014). But no such violation occurs when the court excludes evidence under a valid evidence rule and the defendant retains other ways to defend himself. See id. at 454; United States v. Kerley, 784 F.3d 327, 342–43 (6th Cir. 2015). Sammons used ample alternative means of presenting his false confession defense. He made his case through cross examination, defense witnesses, his own trial testimony, closing arguments, and a jury instruction cautioning that “people, for various reasons, may admit to crimes they did not in fact commit.” R.160 at 84.
C.
Sufficiency of the evidence. “[N]o one may be convicted of a crime based solely on his uncorroborated confession.” United States v. Brown, 617 F.3d 857, 860 (6th Cir. 2010). But if substantial independent evidence corroborates any part of a confession, the entire confession may be used to establish the elements of
Plenty of evidence corroborated Sammons’ confession. Sammons confessed to: (1) sending and receiving child pornography; (2) sending picturеs of the victim to other users; (3) regularly babysitting her; (4) chatting with the personas adopted by Agent Hurst and Officer Peachey; and (5) participating in a chat group featuring discussion of incestuous sexual abuse. Sammons also shared his chat username and provided his phone password to Officer Peachey. The government corroborated each of Sammons’ admissions through a combination of chat records, analysis of his cell phone, printouts of his chats with Agent Hurst and Officer Peachey, the uncontested testimony of trial witnesses, and his own admissions at trial. This extensive corroboration amply supports the confession‘s reliability.
Establishing the reliability of Sammons’ confession all but disposes of his evidentiary challenge. In assessing a sufficiency challenge, we ask whether, after construing all evidence in favor of the verdict, “any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt.” Jackson v. Virginia, 443 U.S. 307, 319 (1979). A defendant who takes a closeup photo of a minor‘s genitals with the intention of producing child pornography is guilty of child sexual exploitation. See
Sammons, to be sure, clings to the contention that he falsely confessed. But the jury considered and rejected that excuse. The confession was amply corroborated and when competing explanations are presented through admissible evidence, we let the jury decide which story to believe. United States v. Wallace, 51 F.4th 177, 182–83 (6th Cir. 2022).
We affirm.