dissenting.
This case basically boils down to the relative import of two Latin phrases: “post *435hoc ergo propter hoc ” and “res ipsa loqui-tur.” The former is a well known logical fallacy (recognized as such since Aristotle’s Rhetoric). It is the fallacy of saying that because effect A happened at some point after alleged cause B, the alleged cause was the actual cause. Such logic has never been enough to survive summary judgment. See, e.g., Abbott v. Federal Forge, 912 F.2d 867, 875 (6th Cir.1990) (“[P]ost hoc, ergo propter hoc is not a rule of legal causation.”).
The latter phrase applies to a narrow class of cases in which the connection between an untoward effect and some type of fault is so clear (and the likelihood of an alternative explanation so low) that no other evidence is required to uphold a jury verdict. The original, and classic, exposition of this principle is Byrne v. Boadle, 159 Eng. Rep. 299 (Exch.1863), a nineteenth century English case where a pedestrian on the streets of Liverpool was struck by a barrel of flour that came flying out of the second-story window of a commercial storeroom. As I will explain below, it seems clear to me that the venerable British case does not describe the case before us. A barrel of flour is extremely unlikely to come flying out onto a city street without some fault by those charged with the care of similar barrels, and it is also extremely unlikely that the flying barrel of flour came from any place other than the adjacent flour warehouse.
In our case, the plaintiffs’ symptoms, which worsened at a later time and after medical care, and which are known to have a wide variety of possible causes, are much less obviously connected to an unspecified dose of a potentially poisonous pesticide. Instead, finding fault, without more, in the latter circumstances represents classic post hoc reasoning. Something potentially causative happened at one time; something untoward happened at a later time. Therefore, plaintiffs allege, the latter must have been caused by the former. The applicable law of Michigan does not permit that unsubstantiated connection to be made, and that is why expert opinion is required. I therefore respectfully dissent from the majority’s invocation of res ipsa in a case where it is not warranted.
I
This case involves assessing the connection between exposure to some chemical substance and a series of physical symptoms that are generally attributable to a wide variety of causes. The district court held that the experts that the plaintiffs relied on to connect the alleged exposure (and the defendant’s behavior relating to it) to the plaintiffs’ symptoms could not testify to any causal relationship. The majority affirms that decision but nevertheless reverses summary judgment, holding that ordinary experience suffices to connect any chemical used (regardless of composition or dose) to the symptoms. In my view, this resolution is not supported by common sense or by the Michigan law that governs the case. Both counsel that a lay juror cannot be expected to understand the complex medical and scientific facts that necessarily underlie any such an attribution of fault and, accordingly, require expert explanation prior to allowing a jury verdict. The majority’s assessment that in this case (and, one supposes, unlike in most toxic tort cases) there is sufficient evidence for the jury to charge plaintiffs’ illnesses to the defendants because the illness began to develop reasonably soon after the exposure — which is only post hoc ergo propter hoc — reinforces rather than refutes this preference for expertise.
To fully understand the weight that the majority asks the post hoc fallacy to bear, it is useful to begin with a point of agreement between the majority and this opin*436ion. The plaintiffs hoped that their treating physicians could testify not just to the physical symptoms with which they were diagnosed but also to the likely cause of such symptoms. That testimony would have concluded that the plaintiffs’ illnesses were explained best by chemical poisoning and that the exposure to whatever pesticides defendants had used was the best explanation of how plaintiffs came in contact with the chemicals that poisoned them. As rehearsed and affirmed in the majority opinion, the district court held that these doctors lacked sufficient expertise to make the causal connection alleged by plaintiffs.
Plaintiffs also proffered the testimony of a second set of experts to demonstrate duty and breach. These liability experts (whose exclusion was not challenged on appeal) would have testified to the dangers of pesticides and an appropriate standard of care for their use. The district court excluded them because they prepared no report and did not plan to give an opinion about the chemicals that defendants actually used or the precautions defendants actually undertook.
Together, this missing expert testimony means that the summary judgment record contains no admissible evidence that directly shows a breach of duty or that shows causation between the alleged breach and the illnesses. To be sure, the plaintiffs’ evidence recounts the spraying of the pesticide and details the potential toxicity at some unspecified dose of the chemicals that probably were sprayed by defendants. But establishing these propositions does not establish how the room should have been sprayed or how the risk of chemical poisoning should have been reduced. That is, the evidence does not tend to prove that the defendants’ activities, in light of the relevant standard of care, were negligent or that the chemicals that were sprayed caused the illness.
Of course, the absence of evidence is not the same as evidence of absence. The majority holds that in the place of the absent evidence there is sufficient circumstantial evidence of breach and causation. Specifically, the historical fact that some illness came quickly on the heels of exposure speaks for itself and makes up for the otherwise missing causal links.
II
Thus presented, the question is whether the plaintiffs needed expert testimony in this case to prove how much chemical exposure is too much chemical exposure or to prove whether the amount of exposure actually caused the alleged harmful consequence. In my view, the majority pays too little attention to this issue, rushing from the fact of exposure and odd symptoms to the legal conclusion of fault. It is of course correct that under Michigan law some complex cases involve breach or causation questions within the ken of the jury notwithstanding the professional or scientific nature of the litigation. See Thomas v. McPherson Health Center, 155 Mich.App. 700, 400 N.W.2d 629, 631 (1986). But the majority assumes, without citation or authority, that this case is such a case.
In Michigan, expert testimony in professional negligence (and toxic tort) cases like this one is required to avoid summary judgment “unless the lack of professional care is so manifest that it would be within the common knowledge and experience of the ordinary layman that the conduct was careless.... ” Lince v. Monson, 363 Mich. 135, 108 N.W.2d 845, 848 (1961). The Michigan courts do not provide a test for what is common knowledge, but do require more than “a bad result,” Jones v. Porretta, 405 N.W.2d 863, 874 (Mich.1987), and have frequently held that negligence can*437not be inferred based on ordinary knowledge simply from an unexpected injury. See, e.g., Woodard v. Custer, 473 Mich. 1, 702 N.W.2d 522, 525 (2005) (“[Wjhether a leg may be fractured in the absence of negligence when placing an arterial line ... in a newborn’s leg is not within the common understanding of the jury....”). This approach comports with the general view that injuries in professional negligence cases, especially those involving complex chemicals and human health, “are usually not immediately obvious and the connection between exposure and injury is not a matter of common sense or everyday experience.” In re Meridia, 328 F.Supp.2d 791, 798 (N.D.Ohio 2004).
As I understand it, these cases require expert testimony in complex, professional, or scientific-based negligence cases in order to limit the dangers associated with indulging the post hoc impulse: it is too easy to charge an uncommon harm to the presence of a mysterious substance. Properly credentialed expert testimony operates as a bulwark against such fallacious attribution of guilt. As in the Daubert context, our concern in applying these cases should be to “assure that the powerful engine of tort liability ... points towards the right substances and does not destroy the wrong ones.” General Electric v. Joiner, 522 U.S. 136, 148-49, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997) (Breyer, J. concurring).
While our case is not a “professional care” case, in a very recent, albeit unpublished, decision, the Michigan Court of Appeals applied this wisdom to a case similar to ours. The court, citing favorably to the district court opinion in this case, held that without expert testimony directly connecting the level of pesticide exposure to the plaintiffs mysterious illness, a plaintiff alleging pesticide poisoning could not get to the jury. See Trice v. Oakland Development Ltd. Partnership, 2008 Mich.App. LEXIS 2484 at *30 (Mich.Ct.App. Dec. 16, 2008) (citing Gass v. Marriott, 501 F.Supp.2d 1011, 1023 (W.D.Mich.2007)). Specifically, “the dose of chemicals to which plaintiff had been exposed had not been determined,” id. at *32, and so “without evidence that plaintiff had been exposed to any chemicals at a level that would be harmful, plaintiff could not establish specific causation.” Id. at *35.
It is against this background that I disagree with the majority’s conclusion that the plaintiffs’ proof, without expert testimony, survives summary judgment.
Ill
As a general matter, the weakness in the majority’s reasoning is demonstrated by reference to the emotive language used to characterize the facts. The words assume negligence and, accordingly, make it easy to agree that a lay person could come to an informed conclusion about the case. For instance, the cloud of pesticide was “toxic or hazardous,” Op. at 431, and it delivered a “high dose,” Op. at 433-34, of chemical exposure because of defendants’ “unacceptable behavior,” Op. at 430-31. But one cannot know that these conclusions (that should be based on scientific facts of how a chemical impacts the human body and legal standards) are appropriate without an expert explaining what amount constitutes a high dose or how much exposure makes a chemical toxic to the human body. Cf. Woodard, 702 N.W.2d at 526-27. It is surely common experience that pesticides are poison, but that does not resolve the question at issue in this litigation: it may be that being exposed to a room “sort of cloudy” with Demand CS will cause no lasting effects if the exposure is five minutes but not ten; ten but not twenty.
The majority avoids the difficulty of scientific judgments by simply defining the *438exposure as a high dose and the defendants’ behavior as unacceptable. Of course a jury can decide for a plaintiff if it is shown that because of a defendant’s action a poisonously high dose of a pesticide was administered to the plaintiff. But a closer examination of the summary judgment record reveals that the evidence the plaintiffs have adduced does not establish anything close to that description of defendants’ behavior and the majority’s holding is premised on a mistaken belief that a lay jury is competent to set the standard of care for the administration of pesticides and to determine the cause of a mystery illness.
A
In reversing the district court’s breach holding, the majority asserts that a “fact-finder is able to weigh and evaluate the evidence based on his or her ordinary experience,” Op. at 15, because the plaintiffs testimony established that the defendants’ action resulted in a “thick, horrid, acrid, putrid” cloud of pesticides in the room. The “because” in the previous sentence, the effect of which is to make it unnecessary for the plaintiffs’ to introduce expert testimony establishing a breach of duty, is not supported by law.
As to the danger of the chemicals (and presumably the duty of care), the evidence cited by the majority is the following: (1) the substance sprayed left the hotel room “sort of cloudy” and (2) the MSDS report for two of the substances possibly used demonstrates that exposure may result in certain symptoms. As to the defendants’ behavior, the majority cites the following evidence: (1) men in masks entered the hotel room and sprayed pesticides despite the presence of suitcases and other indicia of occupancy and (2) defendants were aware that “some of the chemicals they routinely use could cause serious illness _” Op. at 431.
This is underwhelming proof of defendants’ alleged breach of a duty of care. The evidence leaves open more questions than it answers. We do not know how harmful the chemicals are or under what circumstances those harms obtain. For instance, how long do the symptoms persist? How much exposure triggers what symptoms? What measures (besides ventilation) can prevent harm? How many parts per million make a room “cloudy”? How much chemical concentration before a “sort of cloudy” room becomes dangerous? Similarly, the evidence is silent about defendants’ behavior in relation to a standard of care. For instance, which chemical was sprayed? Is a hotel room “well ventilated”? Does the chemical effect dissipate? How fast? What is its effect on articles in a room? How long does it last? Must a room be vacant to be sprayed?
As I understand Michigan’s tort law, the gaps in the evidence suggested by these questions are too wide to be bridged by jury inference. To be sure, the difference between “common knowledge” and a fact that must be explained by expert testimony has not been precisely defined. But wherever the line may be, the questions posed in the previous paragraph about health effects and proper precautions to mitigate them appear to me well beyond the ordinary ken of a juror. See Thomas, 400 N.W.2d at 630 (upholding a directed verdict for the defendant where “[pjlain-tiffs provided expert testimony that [their proffered theories of liability] would constitute a breach of the appropriate standard of care” but “they did not produce evidence in the form of expert opinion that the health center had in fact breached the standard of care.”).
The majority makes no attempt to argue to the contrary based on the plaintiffs’ *439evidence. They assert only that “[e]xpert testimony is not necessary to establish that such egregious behavior does not conform to the standard of care” because “an ordinary person understands that it is unacceptable to enter a place where another is residing and fill that place with airborne poison, without providing for evacuation of the inhabitants, appropriate ventilation, or taking other precautions.” Op. at 430. This assertion fails for two reasons.
First, there is no evidence (lay or expert) to support the majority’s premises about what the defendants did. There is no evidence as to which pesticide was used (that is, how poisonous the “airborne poison” actually was — on the plaintiffs’ allegations, there are differences between Demand CS and Suspend SC); that it would linger long enough to seriously harm someone; that complete evacuation of all property in the room was necessary; that a modern hotel room is not appropriately ventilated; or that “other precautions” were necessary. The assertion thus fails on its own terms.
Second, it asserts without argument that the standard of care is common knowledge. This is unsupportable. Some pesticides can be used in a home by a private individual without supervision or extra preparation and some require the tenting and total evacuation of the home. Simply reading the MSDS reports (whose warnings as to Suspend SC and Demand CS mirror the warnings on the can of over-the-counter Ant & Roach Killer in my chambers1) as the majority does cannot explain the difference. The majority may be correct that if the jury knew all that was required of the defendants under the appropriate level of care, it could compare the plaintiffs’ version of events to that duty and make a breach determination. But that is not the case we have before us. In our case an expert is required to explain the potential hazards associated with certain pesticides and the best practices for avoiding those hazards.
B
The majority’s causation analysis is even less persuasive. It boils down to an assertion that there is evidence of causation because the defendants sprayed a pesticide, scientific evidence shows that one of the pesticides they may have sprayed causes certain physical symptoms, and the plaintiffs did in fact experience those symptoms. The missing premises from this argument — -that defendants did spray the substance known to be dangerous, that the spray was in sufficient amount to cause harm, that plaintiffs’ minutes-long exposure was sufficient to cause harm, that other causes can be excluded with confidence, etc. — are almost too many to list.
The majority holds that those premises can be supplied by inference. The opinion, however, makes no attempt to explain why the causal link between the defendants’ actions and the plaintiffs’ illnesses are within common experience. There is good reason to think that it is not. For instance, ordinary understanding of everyday medical problems does not include the proposition that black tongue is ordinarily caused by spraying of pesticides.2 To be *440sure, an ordinary lay person probably begins with an assumption that black tongue is evidence of something gone wrong, but the question here is what that something is and whether it is chargeable to the defendants’ actions. See Thomas, 400 N.W.2d at 631 (rejecting an argument similar to plaintiffs’ because the “injury was susceptible to a number of explanations, all of which required medical knowledge to discern.”).
The majority’s reliance, at pages 20-21, on Dr. Natzke’s statement that pesticide exposure could cause black tongue to show that the jury would possess the knowledge necessary to make a reliable attribution of fault is not persuasive. First, Dr. Natzke does not say that the ‘plaintiffs’ injuries are the result of pesticides or even that these pesticides could cause black tongue. Second, not only is there a lack of knowledge about how the condition comes about, reliance on Natzke’s statement would permit lay people to make a determination about the cause of an unfamiliar medical condition based only on a post hoc temporal connection and an abstract statement of a risk of harm.
IV
We need to look no further than this case for an illustration of the concerns underlying my belief that these standard of care and causation issues require expert explanation. The flaws in the majority’s reasoning — eliding the difficult scientific questions; conflating colloquial usage of terms like “toxic” and “high dose” with scientific conclusions about the health effects of the plaintiffs’ exposure; and attributing causation on the basis of order of events — are the reasons lay people (jurors and judges alike) are advised to take expert guidance in drawing scientific conclusions. I believe our courts should require that guidance. I respectfully dissent.
. For example, compare the warning the majority cites for the proposition “mere skin contact with Demand CS is dangerous” at page 15 of the opinion with the over-the-counter warning regarding contact with skin or clothing: "Take off contaminated clothing. Rinse skin immediately ... for 15-20 minutes. Call a poison control center or doctor for treatment advice.”
. Indeed, publically available information suggests many other, more common, causes: (1) changes in the normal bacteria or yeast content of the mouth following antibiotic treatment; (2) poor oral hygiene; (3) medi*440cations containing bismuth, such as Pepto-Bismol; (4) regular use of mouthwash containing oxidizing agents; and (5) drinking excessive amounts of coffee or tea. See Alan Carr, What Causes a Black Hairy Tongue? Mayo Clinic: Ask a Dental Specialist, available at http://www.mayoclinic.com/health/ black-hairy-tongue/HQ00325. Similarly, an OSHA document included by plaintiffs in the summary judgment record that describes their diagnosis of “Multiple Chemical Sensitivities” admits that "[tjhere is insufficient scientific evidence to confirm a relationship between ... possible causes and symptoms.” J.A. 564.