(dissenting). I share the majority’s lament that had the trial judge submitted this case to the jury, especially after a long and bitterly contested trial, scarce judicial resources would have been spared given the majority’s view of the matter. Even so, I believe that the trial judge, having presided over the trial and heard all of the testimony from the lips of the witnesses, correctly concluded that the plaintiff failed to establish a causal link between the various adverse employment actions and the plaintiff’s support of his wife’s ill-fated sexual harassment claim.1
We have emphasized that the presence of retaliation is largely *49a question of fact; however, “in a matter such as this the very concept of unlawful retaliation contains significant legal elements, and it may not be simply relegated to jury determination as a purely factual matter without guidance or definition.” Bain v. Springfield, 424 Mass. 758, 765 (1997).
The majority emphasizes that the plaintiff was a stellar performer on the faculty of the biochemistry and molecular biology department (department) of the University of Massachusetts Medical Center (medical center) prior to supporting his wife’s sexual discrimination claim in 1991 and that, as a result, his standing plummeted. From that circumstance, the majority concludes that a jury question arose with respect to retaliatory causation. A close reading of the record suggests otherwise. In the late 1980’s, the medical center’s Scientific Council formed a subcommittee to evaluate the plaintiff’s Protein Chemistry Core Facility (laboratory). Professor Alan Jacobson, who served on the council, testified that the subcommittee was concerned with the quality of work performed at the laboratory and elaborated on the reasons for that negative report. The bottom line was that a recommendation was made to replace the plaintiff with a new director rather than continue under his stewardship. That recommendation was made to the Scientific Council by its subcommittee on February 8, 1990.
Twenty days after learning of the recommendation of the Scientific Council’s subcommittee, which combined the plaintiff’s laboratory with another laboratory, the plaintiff’s wife approached the medical center’s equal employment officer with her complaint regarding sexual harassment. The other funding unit for the laboratory, the Diabetes and Endocrinology Research Center, had concerns similar to those of the subcommittee prior to January, 1991, and based upon those preexistent complaints about the plaintiff’s leadership, voted to withdraw funding from his laboratory in May, 1991.
With respect to the plaintiff’s teaching function, following the academic year 1989-1990, the defendant Frank Chlapowski, as interim department chair, removed the plaintiff from teaching duties based on negative reports called to his attention by *50Professor Reid Gilmore, the graduate course coordinator. Gilmore informed them, and testified at trial, that the plaintiff was ranked “fair to poor” by the students, whereas other teachers in the course were rated “good to excellent.”
The record evidence also shows that members of the department did not want the plaintiff representing them because he had alienated members of the department through his conduct — at one point verbally attacking Chlapowski at a December, 1990, department meeting and asserting that he was not capable of chairing the department.
All of this evidence stands in contrast to the majority’s view that the various adverse acts happened subsequent to engaging in protected activity. I realize that the question before us in reviewing the Superior Court’s ruling is whether there exists anywhere in the evidence, from whatever source derived, any combination of circumstances from which a reasonable inference could be drawn in favor of the plaintiff. Rolanti v. Boston Edison Corp., 33 Mass. App. Ct. 516, 520 (1992). As noted by the United States Court of Appeals for the First Circuit in Mesnick v. General Elec. Co., 950 F.2d 816, 824-825 (1st Cir. 1991), cert. denied, 504 U.S. 985 (1992), in the discrimination context, courts confronted by summary judgment motions must at this point focus on the ultimate question, scrapping the burden shifting framework in favor of considering the evidence as a whole. See also Williams v. Cerberonics, Inc., 871 F.2d 452, 458-459 (4th Cir. 1989). The problem in the instant case, however, is that there is no evidence in the record, other than what the majority characterizes as a “spiral downward,” ante at 44, which specifies the plaintiff’s ultimate burden of proof of retaliation. “Whatever the sources of his proof, a plaintiff, in order to survive judgment as a matter of law, must present evidence from which a reasonable jury could infer that the employer retaliated against him for engaging in [protected] activity.” Mesnick v. General Elec. Co., 950 F.2d at 828. For instance, in Sattar v. Motorola, Inc., 138 F.3d 1164, 1170 (7th Cir. 1998), the court agreed that the plaintiff established its initial burden, but failed to meet its ultimate burden where, as here, there was reliance on a “cat’s paw” theory of liability. If *51one takes into account the uncontested evidence that the plaintiff’s performance was unsatisfactory prior to his engaging in the protected activity, the inferential basis for the majority’s view evaporates.
The plaintiffs claim also fails because he offered no evidence to support his theory of retaliation other than the fact that defendants Michael Czech and Chlapowski knew about his wife’s complaint before their alleged adverse actions against him. The coincidence of timing or “[t]he mere fact that one event followed another is not sufficient to make out a causal link.” MacCormack v. Boston Edison Co., 423 Mass. 652, 662 n.11 (1996). See Hoeppner v. Crotched Mountain Rehabilitation Ctr., Inc., 31 F.3d 9, 14-15 (1st Cir. 1994) (conclusory allegation of retaliation unsupported by specific facts insufficient).2
At a bench conference on the last day of trial, the judge pressed plaintiffs counsel to cite evidence linking the plaintiffs problems to the sexual harassment complaint. Counsel replied, “I think that’s a reasonable inference to draw because nothing happened for nine years until he found out about it. That’s the only way I can do it is by reasonable inference. . . . All I can say is it was going fine and then something happened. And the something that happened was [the] sexual harassment complaint, because there is nothing else.” This is simply not enough. See Lewis v. Gillette Co., 22 F.3d 22, 25 & n.2 (1st Cir. 1994). If knowledge of a discrimination claim preceding an adverse action, itself, could support an inference of causation, then all cases in which employers knew about the discrimination claim would survive a directed verdict challenge based on causation. *52They do not. See, e.g., McMillan v. Massachusetts Soc. for the Prevention of Cruelty to Animals, 140 F.3d 288, 309 (1st Cir. 1998), cert. denied, 525 U.S. 1104 (1999). In addition, “[w]ere the rule otherwise, then a disgruntled employee, no matter how poor his performance or how contemptuous his attitude toward his supervisors, could effectively inhibit a well-deserved discharge by merely filing, or threatening to file, a discrimination complaint.” Mesnick v. General Elec. Co., 950 F.2d at 828.
The record does not warrant an inference of connection between the adverse employment actions and the plaintiff’s support of his wife’s complaint. Disregarding the defendants’ evidence (that the adverse actions were taken solely in response to the plaintiff’s startlingly poor work ethic) leaves a record bereft of any evidence on the subject of causation at all. The plaintiff had an opportunity to cross-examine two members of the ad hoc committee, Professors Thomas Miller and Reid Gilmore, regarding their decision-making process but elected not to do so. Rather, the plaintiff attempted to establish their animus towards him, based on the fact that Miller and the plaintiff previously had clashed over the plaintiff’s “wasting [Miller’s] laboratory staff’s time” asking them questions about another grievance that the plaintiff had filed, and on the fact that Gilmore previously had signed a November 8, 1993, letter with eight other faculty members seeking the plaintiff’s removal for being “deliberately unproductive” and for misrepresenting his publication record. Even assuming a jury believed that Miller and Gilmore disliked the plaintiff on these accounts, there was no testimony or other evidence linking their recommendations to the plaintiff’s support of his wife’s claim.
The evidence also failed to establish any connection between the protected activity and medical center Chancellor Aaron Lazare’s recommendation to the University of Massachusetts’s board of trustees that the plaintiff be terminated, eight years later in May, 1998. Lazare testified he made the recommendation that the plaintiff be terminated because he was performing no work, and his contentious nature hampered Lazare’s at*53tempts to recruit a chairperson for the department.3 As with the testimony of Gilmore and Miller, Lazare’s testimony provided no evidence that he considered the plaintiff’s support of his wife’s harassment claim when Lazare recommended the plaintiff’s discharge.
Despite the fact that Lazare was the sole witness at trial who addressed his discharge, the plaintiff asked Lazare only one question — whether there was a medical center procedure to insure that there would be no retaliation when a low-level employee makes a complaint about her supervisor. The plaintiff failed to refute or even to explore the facts and circumstances that led to Lazare’s recommendation to discharge him. Once again, there is simply no evidence linking Lazare’s recommendation to the plaintiff’s support of his wife’s claim.
Two learned commentators have phrased it as follows: “Comments by coworkers, sworn statements from the plaintiff, and the timing of the adverse employment action in relation to the protected activity also may be introduced as circumstantial evidence of retaliation. However, whatever type of evidence is admitted, it is crucial that the evidence be linked back to the ultimate decisionmaker and, in particular, to the adverse employment decision that occurred. Some vague notion of workplace misconduct disconnected from the adverse action is not probative of retaliation’'’ (emphasis added). Snell & Eskow, What Motivates the Ultimate Decisionmaker? An Analysis of Legal Standards for Proving Causation and Malice in Employment Retaliation Suits, 50 Baylor L. Rev. 381, 396 (1998).
The plaintiff “tendered nothing, direct or circumstantial, suggesting a retaliatory animus. To the contrary, the record, read as a whole, is more consistent with an employer’s longstanding desire to improve an employee’s behavior than with some sort of vengeful preoccupation.” Mesnick v. General Elec. Co., 950 F.2d at 828.
I would affirm the trial judge’s directed verdicts on the plaintiff’s claims against all four defendants.
The plaintiff’s wife subsequently pursued her claims in court. A jury ultimately found in favor of the defendants with respect to her sexual harass*49ment claims. This court affirmed the judgment. See Anderson-Mole v. University of Mass., 49 Mass. App. Ct. 723 (2000).
The majority skirts the temporal issue by characterizing the first negative evaluation that is not time-barred, in February, 1993, as close in time to the plaintiffs wife’s sexual harassment filing in January, 1991.1 read the cases on this point differently. See Cooper v. North Olmsted, 795 F.2d 1265, 1272 (6th Cir. 1986) (mere fact that bus driver was discharged four months after she filed discrimination claim was insufficient to support inference of retaliation); Figgous v. Allied/Bendix Corp., Allied-Signal, 906 F.2d 360, 362 (8th Cir. 1990) (discharge of employee approximately one year after he filed administrative charge of discrimination was too remote in time to create inference of retaliatory motive); Morgan v. Musselwhite, 101 N.C. App. 390, 393 (1991) (holding that two-year delay between plaintiffs filing of worker’s compensation claim and subsequent termination from employment was insufficient to establish prima facie case of retaliation).
The plaintiff himself testified that he had brought grievances, lawsuits, or charges of misconduct against no less than nine different faculty members in his nine or ten years with the medical center.