AMENDED OPINION*
The oenologist’s creed teaches that we should drink no wine before its time. Much the same principle applies to summary judgment; it is a deliciously helpful device if properly timed, but one that can leave a sour taste if brought to bear on an insufficiently fermented record.
This appeal, which follows on the heels of a substantial verdict in a defamation suit, is a paradigmatic example of the point. The briefs before us raise a plethora of First Amendment issues. At the threshold, however, lies the preliminary question on which the appeal ultimately hinges: did the district court, ruling on a pretrial motion, appropriately determine that the plaintiff, a Maryland assistant state’s attorney, was a private figure and not a public official for libel-law purposes? Concluding, as we do, that the court’s ruling was premature, we answer that question in the negative. Accordingly, we vacate the judgment below and remand for a new trial.
I. THE LEGAL LANDSCAPE
In order to put this appeal into perspective, we think it is useful to begin by rehearsing a salient aspect of the substantive law of defamation: the public official/private figure dichotomy.
It is apodictic that the First Amendment “prohibits a public official from recovering damages for a defamatory falsehood relating to his official conduct unless he proves that the statement was made with ‘actual malice’ — that is, with knowledge that it was false or with reckless disregard of whether it was false or not.” New York Times Co. v. Sullivan, 376 U.S. 254, 279-80, 84 S.Ct. 710, 11 L.Ed.2d 686 (1964). The New York Times rule extends not only to public officials but also to public figures, whether all-purpose or limited-purpose. See Gertz v. Robert Welch, Inc., 418 U.S. 323, 345, 94 S.Ct. 2997, 41 L.Ed.2d 789 (1974).
The distinction between public figures and public officials, on the one hand, and purely private figures, on the other hand, has potentially profound consequences in a defamation case. Generally speaking, the status that the plaintiff occupies along the public/private continuum will determine what he must prove in order to recover damages. Leaving to one side the imposition of liability without fault (as to which certain restrictions pertain), states may shape their own standards of liability when a defamation action involves a private-figure plaintiff. See id. at 347, 94 S.Ct. 2997. Under Massachusetts law, for example, that standard is negligence. See Stone v. Essex County Newspapers, Inc., 367 Mass. 849, 330 N.E.2d 161, 168 (1975). That is far less demanding, from the plaintiffs standpoint, than the “actual malice” standard that obtains when the plaintiff is a public official or public figure. See New York Times, 376 U.S. at 279-80, 84 S.Ct. 710. A plaintiffs status, therefore, shapes the course of any defamation litigation. If he is a public official or public figure, he must prove actual malice with “convincing clarity.” Id. at 285-86, 84 S.Ct. 710. If, *202however, he is a purely private figure, it suffices (at least in Massachusetts) to prove negligence by a preponderance of evidence. See Stone, 330 N.E.2d at 174-75.
In defamation cases, public-figure status has the same legal ramifications as public-official status — but the two terms are not synonymous. See generally Kassel v. Gannett Co., 875 F.2d 935, 941 n. 4 (1st Cir.1989). Public officials, as the term implies, are those who hold particular kinds of public office. See, e.g., Time, Inc. v. Pape, 401 U.S. 279, 284, 91 S.Ct. 633, 28 L.Ed.2d 45 (1971) (deputy chief of detectives); New York Times, 376 U.S. at 283 n. 23, 84 S.Ct. 710 (elected city commissioner). Public figures may or may not be public officials; they are persons who “have assumed roles of especial prominence in the affairs of society.” Gertz, 418 U.S. at 345, 94 S.Ct. 2997. Commonly, “those classified as public figures have thrust themselves to the forefront of particular public controversies in order to influence the resolution of the issues involved.” Id. For present purposes, we need not dwell either on the distinction between public officials and public figures or on the exact dimensions of public— figure status; the defendants do not contend that the plaintiff is a public figure— only a public official.
The public-official classification eludes precise definition. See Kassel, 875 F.2d at 939. Not every public employee is a public official for libel-law purposes. See Hutchinson v. Proxmire, 443 U.S. 111, 119 n. 8, 99 S.Ct. 2675, 61 L.Ed.2d 411 (1979). To the contrary, only public employees with “substantial responsibility for or control over the conduct of governmental affairs” should be deemed public officials for libel-law purposes. Rosenblatt v. Baer, 383 U.S. 75, 85, 86 S.Ct. 669, 15 L.Ed.2d 597 (1966). Moving from the general to the particular is, however, a daunting task; it is difficult to tell “how far down into the lower ranks of government employees” the public-official designation extends. New York Times, 376 U.S. at 283 n. 23, 84 S.Ct. 710.
II. FACTS AND PROCEEDINGS BELOW
In its January 10-16, 2003 issue, The Boston Phoenix, a weekly newspaper, published a “special report” authored by Kristen Lombardi. Written as an exercise in investigative journalism, the piece ran for nine pages under the title “Children at Risk.” Its central thesis bemoaned what the reporter had determined to be an apparent trend in family courts: that when a mother accuses a father of child abuse in a child custody dispute, those courts, ill-equipped to handle such charges, often award full custody to the father. The article reviewed three scientific studies of custody-dispute outcomes and recounted the personal experiences of four families enmeshed in the system.
One such case history chronicled a custody clash between Sarah Fitzpatrick and Marc E. Mandel (who was, at the time, an assistant state’s attorney in Maryland). That case history appeared under the subheading “Losing custody to a child molester.” The story recounted the sordid battle waged by the couple over custody of their two minor children, A.R.M. and J.P.M. (pseudonymously referred to by Lombardi as “Amy” and “James”), and dwelt in some detail on Fitzpatrick’s allegation that Man-del was a child molester. For example, it reported Fitzpatrick’s suspicions about Mandel’s relationship with J.P.M. and gave prominent play to a Baltimore County Department of Social Services (DSS) investigation into allegations that Mandel had abused his daughter from a previous marriage, A.N.M. In that regard, the article *203related that a “report conducted for the Baltimore County DSS determined that Mandel had assaulted” A.N.M. The article went on to state that despite the claims of abuse, a Maryland family court judge awarded Mandel full custody of A.R.M. and J.P.M. and denied Fitzpatrick any visitation rights, labeling her “a pathological liar, or a purposeful liar, or both.”
After The Phoenix published the article in print and on the internet, Mandel began receiving negative work evaluations. He later left his post and became self-employed in the private practice of law.
Invoking diversity jurisdiction, see 28 U.S.C. § 1332(a), Mandel filed suit for defamation against, inter alia, The Phoenix, its various corporate personas, two of its editorial gurus (Peter Kadzis and Susan Ryan-Vollmar), and Lombardi (collectively, the defendants) in the United States District Court for the District of Massachusetts. Shortly after answering the complaint, the defendants moved for judgment on the pleadings or, in the alternative, summary judgment. Mandel cross-moved for partial summary judgment. The district court denied both motions, ordered discovery to be completed within three months, and offered to entertain renewed motions at that juncture.
The parties completed discovery and cross-moved for summary judgment. Following a hearing, the court denied the defendants’ motion and granted Mandel’s motion for partial summary judgment on the public-official issue. Mandel v. Boston Phoenix, Inc., 322 F.Supp.2d 39, 44 (D.Mass.2004). Reasoning that Mandel was only a low-level prosecutor who “did ordinary legal work,” id. at 42, did not interact with the press, and did not assume a risk of diminished privacy through his employment, the court concluded that he was not a public official within the purview of the libel laws, see id. at 42-43.
The litigation morphed into a ten-day jury trial. Once Mandel had presented his case in chief, the defendants moved for judgment as a matter of law. See Fed. R.Civ.P. 50(a). The court granted that motion in part, concluding that Mandel had not adequately established actual malice.1 At the close of all the evidence, the defendants moved again for judgment as a matter of law. See id. The court denied that motion.
The court submitted the case to the jury, which found for Mandel and awarded him $950,000 in damages. As indicated on a special verdict form, it found two statements in the article actionable: (i) the subheading “Losing custody to a child molester” and (ii) the comment that “a July 2002 report conducted for the Baltimore County DSS determined that Mandel had assaulted his 10-year-old daughter from an earlier marriage. The report states that Mandel’s daughter accused him of seven ‘incidents’ of fondling....”
In the aftermath of the trial, the defendants moved to alter or amend the judgment, see Fed.R.Civ.P. 59(e), on the ground that evidence adduced at trial and not previously available demonstrated beyond hope of contradiction that Mandel was a public official within the purview of the libel laws. The defendants also moved for judgment as a matter of law, see Fed. R.Civ.P. 50(b), claiming that Mandel’s evidence was insufficient to establish (i) the falsity of the two statements, (ii) negligence in their publication, or (iii) any in*204cremental harm. The court denied both motions. This timeous appeal followed.
III. ANALYSIS
At the epicenter of this appeal lies the district court’s status determination. We start there.2 Finding reversible error in the timing of that determination, we abjure review of the vast majority of the other issues raised by the defendants. The lone exception is the district court’s refusal to grant the motion for judgment as a matter of law. That decision still must be scrutinized; after all, if Mandel did not present sufficient evidence of the elements of a private-figure defamation claim, the case would be over (as he certainly could not meet the heightened burdens associated with public-official status).
A. The Status Determination.
Typically, courts approach the public-official analysis as if it were a three-legged stool, taking into account: (i) the extent to which the inherent attributes of a position define it as one of influence over issues of public importance; (ii) the position’s special access to the media as a means of self-help; and (iii) the risk of diminished privacy assumed upon taking the position. See Kassel, 875 F.2d at 939-40. Here, however, we focus on a procedural milestone — the district court’s pretrial determination that the record was sufficiently developed to permit it to decide whether Mandel was — or was not — a public official.
Given the significance of status determinations in libel-law litigation, it is often perfectly reasonable to attempt to decide whether a plaintiff is a public official or public figure during pretrial proceedings. See, e.g., Nicholson v. Promotors on Listings, 159 F.R.D. 343, 343-45 (D.Mass.1994); see also Miller v. Transam. Press, Inc., 621 F.2d 721, 724 (5th Cir.1980) (advising that the question of public-figure status “be answered as soon as possible”). But although status determinations normally are “grist for the court’s — not the jury’s — mill,” Pendleton v. City of Haverhill, 156 F.3d 57, 67 (1st Cir.1998), such questions are inescapably fact-specific, see id. at 70 (describing public-figure conclusion as “factbound”); Penobscot Indian Nation v. Key Bank, 112 F.3d 538, 561 (1st Cir.1997) (acknowledging that “a finding of public figure status necessitates a detailed fact-sensitive determination”).
In other words, a plaintiffs status, though calling for a legal determination, is heavily dependent on the underlying factual record. Consequently, there are cases in which “it may not be possible to resolve the [public-figure] issue until trial.” Miller, 621 F.2d at 724; cf. Pendleton, 156 F.3d at 62, 68 (upholding district court’s ruling, made at the conclusion of the evidence, that the plaintiff was a limited-purpose public figure).
It is common ground that appellate review of an order granting summary judgment is confined to the record before the district court at the time it made the challenged ruling. See Voutour v. Vitale, 761 F.2d 812, 817 (1st Cir.1985) (per curiam). Such review is de novo and engen*205ders consideration of the record as then constituted “and all reasonable inferences therefrom in the light most hospitable to the summary judgment loser.” Houlton Citizens’ Coal. v. Town of Houlton, 175 F.3d 178, 184 (1st Cir.1999). When the facts, so marshalled, show “that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law,” summary judgment is generally thought to be appropriate. Fed.R.Civ.P. 56(c); see DePoutot v. Raffaelly, 424 F.3d 112, 117 (1st Cir.2005). The presence of cross-motions for summary judgment neither dilutes nor distorts this standard of review. See Alliance of Auto. Mfrs. v. Gwadosky, 430 F.3d 30, 34 (1st Cir.2005); Blackie v. Maine, 75 F.3d 716, 721 (1st Cir.1996).
As noted above, summary judgment requires that the moving party exhibit an entitlement to a judgment as a matter of law. See Fed.R.Civ.P. 56(c). That requirement is not a hollow one. Summary judgment “should be granted only where ... [further] inquiry into the facts is not desirable to clarify the application of the law.” Stevens v. Howard D. Johnson Co., 181 F.2d 390, 394 (4th Cir.1950). In the case at bar, the factual record, at the summary judgment stage, was too uncertain to warrant a legal conclusion either way about Mandel’s status. We explain briefly.
To begin, the factual record was disturbingly thin. Mandel’s summary judgment motion relied almost exclusively on facts derived from three sources: his own affidavit, his deposition testimony, and the deposition testimony of his supervisor, Kim Detrick. These materials indicated that, as an assistant state’s attorney, Mandel received only a modest salary, had little supervisory authority, neither created nor directed policy, and made no decisions that significantly affected government operations. He prosecuted only minor crimes in Maryland’s lowest-tier court and participated in only one jury trial (assisting the lead attorney). As a matter of practice, he directed all press inquiries to a supervisor.
The defendants also made some modest contributions to the summary judgment record. They culled a few additional facts from the Mandel and Detrick depositions and supplemented them with excerpts from the deposition of another supervisor (Sue Schenning), a letter from a vehicular accident victim whose case Mandel had handled, and Mandel’s employment file. This proffer painted a different, though not altogether inconsistent, picture: as an assistant state’s attorney, Mandel handled crimes with penalties of imprisonment, fourth-degree sexual offenses, and violations of protective orders. He conducted “pray jury” trials at the state circuit court level, second-chaired a murder trial (in which he gave the opening statement and conducted some direct examination), nolle prossed a vehicular accident case, and represented to a court that a witness could not testify even though she was present. He interviewed victims of crimes, discussed eases with other attorneys, and had access to the media. His employment file indicated that he was “a prosecutor for [the] State of Maryland” and “a sworn law enforcement [officer] or person whose principal responsibilities are unique to the criminal justice system.”
This factual record, on which the district court based its status determination, contains several gaps, and what facts there are give rise to conflicting inferences. Perhaps most notably, neither side seems to have focused on the question of whether Mandel’s experience as an assistant state’s attorney was a reflection of the attributes inherent in the position or, alternatively, whether he simply was given less responsi*206bility because of his particular proclivities. This is a crucial distinction in the public-official calculus. See Kassel, 875 F.2d at 935 (recognizing that “[t]he inherent attributes of the position, not the occurrence of random events, must signify the line of demarcation” between public-official and private-figure status). Moreover, the record, as it stood, cried out for a special sort of judgment call: which job characterizations — Mandel’s, given in anticipation of litigation, or those contained in his employment file — more accurately depicted the inherent attributes of the position. That judgment call could not be made under the constraints of Rule 56. See Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986) (explaining that “[cjredibility determinations, the weighing of the evidence, and the drawing of legitimate inferences from the facts” are not functions to be performed by a judge on summary judgment). At trial, however, the district court, as the arbiter of Man-del’s status, could have resolved those uncertainties. Cf. Pendleton, 156 F.3d at 62, 68; Miller, 621 F.2d at 724.
Finally, better evidence was readily obtainable. There was a high likelihood that a full-dress trial (or, perhaps, even more pointed discovery) would lead to materially improved development of the record and, thus, clarify the application of the law to the facts. Indeed, at the motion hearing Mandel’s counsel candidly admitted that the question was “so fact-specific” that the status determination could only be made after the court and the jury had “all the evidence.”3
The testimony actually adduced at trial confirms the accuracy of this prediction. For instance, Mandel testified that his reputation was important to him professionally because he was “dealing with the public on a daily basis ... prosecuting criminal cases.” Similarly, Schenning offered insights not apparent from her deposition, especially her trial testimony that assistant state’s attorneys “are the face of [the] State’s Attorney,” with “tremendous power, even at a very early point in their career[,] ... making decisions whether to [nolle prosse] a case, dismiss a case, ... go forward with a case, [and] what kind of sentence to ask for.” The fact that this testimony came to light even though the status determination had been made prior to trial is telling; with the issue on the table, it seems evident that even greater clarification of an assistant state’s attorney’s role would have emerged.
It is also significant in our situation that neither this court nor the Massachusetts appellate courts have addressed the question of whether an assistant state’s attorney (or someone in a comparable position) is a public official for First Amendment purposes. What little case law there is suggests that such a person might be a public official, see, e.g., Henry v. Collins, 380 U.S. 356, 356-57, 85 S.Ct. 992, 13 L.Ed.2d 892 (1965) (per curiam) (treating a “county attorney” as a public official, albeit without discussion of the status determination); Murray v. Bailey, 613 F.Supp. 1276, 1280 (N.D.Cal.1985) (treating a San Francisco assistant district attorney as a public official without discussion of the status determination), but in any event, the inquiry is too fact-dependent to rely exclusively on labels. Further factual development would have allowed for a more precise, more nuanced application of the law.
We recognize that a showing of inadequate factual support for a claim or *207defense sometimes may be enough to warrant summary judgment. See Celotex Corp. v. Catrett, 477 U.S. 317, 325, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986); In re Varrasso, 37 F.3d 760, 763 n. 1 (1st Cir.1994). That doctrine is inapplicable here: the parties failed sufficiently to develop the record, and the evidence that they managed to adduce was inherently imprecise. Still, that evidence, though thin, pointed in different directions; that is, it tended to support conflicting inferences. Summary judgment cannot be predicated on so vacillatory a record. See Varrasso, 37 F.3d at 764 (concluding that when a court must choose between competing plausible inferences, “[that] choice cannot be made under the banner of summary judgment”).
To be sure, the public-official question is ultimately a question of law. We have no doubt that, with sufficient evidence before it, a court — especially a court freed from the constraints attendant to summary judgment — will be capable of making the public-official/private-figure determination. See Pendleton, 156 F.3d at 67. This determination, however, is necessarily a fact-bound one, see id. at 70, and the district court in this case made its ruling under the strictures of the summary judgment standard. Given those impediments, we think that the district court’s resolution of the status question was premature.
We hold, therefore, that the district court should not have essayed a definitive status determination on the exiguous information available to it at summary judgment but, rather, should have demanded more detailed factual development (even if that meant deferring the status determination to the time of trial). Cf. Askew v. Hargrave, 401 U.S. 476, 478-79, 91 S.Ct. 856, 28 L.Ed.2d 196 (1971) (per curiam) (observing that pleadings and an affidavit provided an inadequate basis for deciding an equal protection claim because they did “not sufficiently present the facts”).
We have pondered whether we should attempt to decide, on the full record (that is, the record including the trial testimony) whether Mandel is a public official or a private figure for First Amendment purposes. On reflection, that is not a satisfactory alternative. Because of the premature summary judgment ruling, the parties had no incentive to offer further evidence at trial on the status question. That some new insights made their way into the trial transcript is more a harbinger of tales yet to be told than a completion of the eviden-tiary presentation.
Nor is there any other basis on which we can deem the error harmless. The premature summary judgment ruling determined a question of enormous importance and cast a wide shadow over the entirety of the litigation. Thus, we have no real choice but to vacate the jury verdict and order a new trial in which the facts bearing upon Mandel’s status can be fully and fairly aired.
B. Sufficiency of the Evidence.
In their Rule 50(b) motion, the defendants challenged the sufficiency of the evidence to support jury findings that: (i) the two statements were false; (ii) the defendants were negligent in publishing them; and (iii) the false statements caused incremental harm (i.e., harm apart from that caused by similar non-actionable statements).
Ordinarily, we review the disposition of a motion for a judgment as a matter of law de novo. See Gibson v. City of Cranston, 37 F.3d 731, 735 (1st Cir.1994). In so doing, “we scrutiniz[e] the proof and the inferences reasonably to be drawn therefrom in the light most hospitable to the nonmovant, refraining entirely from differential factfinding,” and “may *208not consider the credibility of witnesses, resolve conflicts in testimony, or evaluate the weight of the evidence.” Cook v. R.I. Dep’t of MHRH, 10 F.3d 17, 21 (1st Cir.1993) (alteration in original) (citations and internal quotation marks omitted). We “can overturn a jury’s verdict and grant judgment in favor of the verdict loser only if the evidence, so viewed, is such that reasonable minds could not help but reach an outcome at odds with the verdict.” Id.
This case, however, is out of the ordinary. Appellate deference to the jury is muted when a case implicates First Amendment principles. See Veilleux v. Nat’l Broad. Co., 206 F.3d 92, 106 (1st Cir.2000). In such an instance, we must undertake independent review of the evidence insofar as it bears on the constitutional issues. See Bose Corp. v. Consumers Union of U.S., Inc., 466 U.S. 485, 508, 104 S.Ct. 1949, 80 L.Ed.2d 502 (1984).
Independent review is not a limitless ransacking of the record as a whole. See Veilleux, 206 F.3d at 107; see also Bose, 466 U.S. at 514 n. 31, 104 S.Ct. 1949 (recognizing that independent review is not equivalent to de novo review of the judgment itself). It does not extend “to all determinations concerning a particular legal claim, but only to those that specifically involve the application of First Amendment law to specific facts.” Veilleux, 206 F.3d at 107. Purely factual determinations (such as credibility calls) remain subject to the usual degree of deference. See id.; see also Eastwood v. Nat’l Enquirer, Inc., 123 F.3d 1249, 1252 (9th Cir.1997) (recognizing that appellate courts undertaking independent review defer “on questions of credibility, which the jury is uniquely qualified to answer”).
In this case, we must independently review whether Mandel satisfied his constitutionally imposed burden of showing the falsity of each of the disputed statements and establishing that the defendants, at the very least, failed to exercise due care in publishing them. See Veilleux, 206 F.3d at 108 & n. 7. But because the question of whether Mandel proved that the statements incrementally harmed him does not implicate the First Amendment, that question remains subject to the conventional standard of review applicable to orders granting or denying Rule 50(b) motions. See Masson v. New Yorker Magazine, Inc., 501 U.S. 496, 523, 111 S.Ct. 2419, 115 L.Ed.2d 447 (1991) (rejecting “any suggestion that the incremental harm doctrine is compelled as a matter of First Amendment protection for speech”); cf. Fiori v. Truck Drivers, Local 170, 354 F.3d 84, 89 (1st Cir.2004) (recognizing that “[wjhether a libel was the but-for cause of subsequent economic loss does not directly implicate the First Amendment, as long as it is clear that the finding was not a hidden award of punitive damages”).
We have undertaken a painstaking perusal of the evidence and find that the verdict passes muster.4 A step-by-step excursion through the record would serve no useful purpose; instead, we forgo such a journey in favor of a synopsis of our conclusions. Cf. DiMillo v. Sheepscot Pilots, Inc., 870 F.2d 746, 750 (1st Cir.1989) (noting that “[t]here is no rule which requires an appellate court to string together facts solely because [litigants] choose[ ] to challenge factbound determinations”).
We begin with the question of falsity. The district court charged the jury (and we assume, for purposes of this analysis) that Mandel was a purely private figure. *209The Constitution requires that, “at least where a newspaper publishes speech of public concern, a private-figure plaintiff cannot recover damages without also showing that the statements at issue are false.” Phila. Newspapers, Inc. v. Hepps, 475 U.S. 767, 768-69, 106 S.Ct. 1558, 89 L.Ed.2d 783 (1986). Mandel cleared this constitutional hurdle with ease, presenting strong evidence that the two statements in question were false (i.e., that he was not “a child molester” and that no “report conducted for the Baltimore County DSS” in July of 2002 had “determined that [he] had assaulted” his daughter by his first marriage).
As to the first statement, Mandel’s proof included medical records, a Massachusetts DSS investigation that failed to uncover any conclusive evidence of abuse as to J.P.M., a specific finding by the Maryland family court judge that Mandel was not a child molester, a Virginia DSS investigation concluding that allegations concerning A.N.M. were unfounded, and Mandel’s trial testimony that he had never inappropriately touched any of his children. As to the second statement, the proof included testimony from a Baltimore County DSS official that the Baltimore County DSS did not retain Dr. Eliana Gil (the author of the July 2002 report) to investigate on its behalf, as well as testimony that the report contained no definitive conclusion that Mandel had molested A.N.M.
In the course of independent review, we are bound by the jury’s credibility determinations. See Eastwood, 123 F.3d at 1252. Given that the jury must have accepted the main thrust of Mandel’s testimony on these points, we conclude that his testimony and the other evidence, in cumulation, established falsity as to both statements.
Private figures also must adduce “proof of negligent publication of a defamatory falsehood.” Stone, 330 N.E.2d at 168; see Veilleux, 206 F.3d at 108 (recognizing a constitutional requirement that “private individuals must prove fault amounting at least to negligence on the part of a media defendant, at least as to matters of public concern”). Put another way, a private-figure plaintiff must show that a media defendant failed “to act reasonably in checking on the truth or falsity ... of the communication before publishing it.” Appleby v. Daily Hampshire Gazette, 395 Mass. 32, 478 N.E.2d 721, 724 (1985) (alteration in original) (citations and internal quotation marks omitted). The question of fault engenders independent review. See Veilleux, 206 F.3d at 108.
We need not tarry. It suffices to say that the reporter, Lombardi, neither read pertinent documents available to her (even though some of them were referenced by others) nor contacted several individuals who might have provided opposing views. She also incorrectly characterized the July 2002 report, misrepresenting its findings and the identity of the party for whom it was prepared.5 Lombardi further admitted that she did not have access to the full Baltimore County DSS file and guessed at its contents. Considering these facts and the strong evidence of falsity at the time the article went to press, an unconditional branding of Mandel as “a child molester” *210may well have been negligent. This is especially so given our obligation, on independent review, to respect the jury’s credibility determinations. See id. at 107; Eastwood, 123 F.3d at 1252.
The defendants’ last point — incremental harm — is a non-starter. The district court charged the jury, at the defendants’ behest, about the need to show incremental harm. No party challenges that instruction and it is not patently wrong; therefore, we treat it as the law of the case. See Milone v. Moceri Family, Inc., 847 F.2d 35, 38-39 (1st Cir.1988).
We owe substantial deference to the jury’s ensuing finding. On this record, there is no sound reason to disavow that deference. For example, the jury rationally may have found the statements that unflinchingly branded Mandel as a child molester especially harmful. Or, alternatively, the jury rationally may have found other, non-actionable statements simply unharmful. Either way, we are bound to conclude that Mandel presented adequate evidence to support the finding of incremental harm.6
IV. CONCLUSION
To recapitulate, we hold that the district court did not err in denying the defendants’ motion for judgment as a matter of law. We nonetheless vacate the judgment due to the court’s premature pretrial decision on the linchpin public-official issue. Summary judgment is proper only when it is appropriately timed — and, given the lack of factual development in the summary judgment record, the timing here was inauspicious. There was no principled way for the district court to hold, at that juncture, that either side was entitled to judgment as a matter of law on the status question.
That leaves the scope of the remand. We recognize that we have wide discretion to remand for a new trial on all, or only some, of the issues in the case. See Dopp v. HTP Corp., 947 F.2d 506, 518 (1st Cir.1991). The touchstone is that a new trial should not “be limited to fewer than all the issues unless it clearly appears that the issues to be retried are so distinct and separable from the other issues that a trial of those issues alone may be had without injustice.” La Plante v. Am. Honda Motor Co., 27 F.3d 731, 738 (1st Cir.1994). In the final analysis, then, the scope of a remand is normally a judgment call for the appellate court. See Wilson v. Mar. Overseas Corp., 150 F.3d 1, 13 (1st Cir.1998).
In this case, there are arguably three categories of issues that are eligible for retrial: liability as to the statements that the jury found actionable, liability as to the statements that the jury found not to be actionable, and damages.7 We are confident that retrial is necessary as to the first and third sets of issues. The status determination is intimately intertwined with the question of liability as to the two actionable statements, and the question of damages is dependent on the liability conclusion.
*211The second category presents a closer call. Arguably, the error that tainted this litigation never influenced the jury’s findings that two statements were not actionable. Yet this case is complex, and it is difficult to determine how the specter of those two statements may have affected other aspects of the trial. In any event, retrial as to these statements will require the parties and the district court to muster only marginally greater resources than otherwise would be the case. We therefore decree that the retrial encompass all the issues.
We need go no further. We caution, however, that a new trial will generate a new record and a new set of credibility determinations. For that reason, our assessment of the evidence adduced at the first trial will not be binding on the lower court or the jury when the case is retried.
Vacated and remanded.