Impediments to dialogue about animal research (part 4).

As we continue our look at ways that attempted dialogues about the use of animals in research run off the rails, let's take up one more kind of substantial disagreement about the facts. Today's featured impediment:

Disagreement about whether animals used in research experience discomfort, distress, pain, or torture.

This disagreement at least points to a patch of common ground shared by the people disagreeing: that it would be a bad thing for animals to suffer. If one party to the discussion accepts the premise that animal suffering is of no consequence, that party won't waste time haggling over how much suffering animal research actually produces.

One extreme position here is that non-human animals cannot "suffer" at all, simply because non-human animals lack the necessary self-awareness to have "feelings" of this sort. Instead, someone on this extreme would regard non-human animals as essentially biochemical machines that display responses to stimuli. If we saw those responses in another human, we might interpret it as suffering, but making such an interpretation of a non-human animal's response would involve ascribing human mental states to this animal.

While this position may have been more commonly voiced back when behaviorism was still in favor, it's not one I've seen seriously advanced in recent discussions. In fact, the animal welfare regulations that govern animal research side-step questions of the animal psyche and self-awareness by focusing on "pain" and "distress rather" than "suffering".

The Institutional Animal Care and Use Committee Guidebook (2nd edition, 2002) helpfully provides definitions of these terms. Pain is "an unpleasant sensory or emotional experience associated with actual of potential tissue damage," and distress is "an aversive state in which the animal is unable to adapt completely to stressors and the resulting stress and shows maladaptive behavior." Under these definitions, neither pain nor distress requires self-awareness. And as we saw in the last post, animal welfare regulations require that researchers avoid or minimize animal discomfort, distress, and pain. (These regulations also require measures to enrich the experimental animals' environment to decrease their boredom.)

The Institutional Animal Care and Use Committee Guidebook also gives some guidance for investigators and IACUCs as far as assessing pain and distress:

Numerous references indicate that both laboratory animals and humans receive and process noxious stimuli using similar mechanisms and with similar thresholds of awareness. The pain tolerance, or maximum stimulus intensity voluntarily accepted, varies between species and between individuals of the same species, including humans. Pain typically results from stimuli that damage tissue or have the potential to damage tissue. An animal's response to pain is often adaptive to reduce movement to minimize reinjury and aid recuperation. However, this response may lead to physiological and behavioral changes which impact negatively on both the animal's well-being and the research results.

Fundamental to the relief of pain is the ability to recognize its clinical signs in various species of animals. Due to the inability of animals to verbalize, it is essential that animal care staff and researchers receive adequate training on how to recognize clinical signs of pain and distress. ...

According to the NRC report on pain and distress, while there are no generally accepted criteria for distress, there are a number of metabolic, physiologic and behavioral parameters that are altered in distressed animals. These include changes in reproductive performance, elevation in glucocorticoid levels and elevation in catecholamine levels. It is necessary to use objective assessments, which means choosing appropriate parameters and quantifying observations. Numerous models for scoring pain and distress have been published and involve assigning a numeric score to observations with the aid of descriptors. It is often useful to start with a general set of observations for assessing pain and distress such as change in body weight, physical appearance/posture or changes in unprovoked and provoked behavior. The assessment system should then be modified on a case-by-case basis using specific changes that may be anticipated in a particular study.

(pp. 110-111)

The upshot here is that you cannot judge the amount of discomfort, distress, or pain which animals experience under a particular experimental protocol from first principles. Rather, you can only make an accurate assessment by observing the animals and interpreting their responses according to an objective standard.

Judgments are made by investigators developing protocols and IACUCs reviewing protocols about what is likely to cause animals pain and distress. These judgments are supposed to be grounded on existing data about what causes pain and distress to the type of animal whose use is proposed (and when in doubt, the presumption is that procedures that would cause pain and distress in humans may also do so in non-human animals). But the investigators performing research under the approved protocol are required to observe the actual animals for signs that may indicate pain or distress, whether it was expected or not, and take the appropriate steps to alleviate it. (Of course, the effectiveness of analgesia, anesthesia, and other steps taken to alleviate pain and distress should also be judged using objective standards.)

Here, if all parties to the attempted dialogue can agree to focus on a particular protocol and a particular objective standard for judging pain and distress and the effectiveness of measures intended to alleviate them, there is a hope of coming to agreement about the amount of animal pain and distress in a discrete piece of research.

Such agreement is not possible, however, if any of the parties accept the objective standards only in the cases where they produce the desired answer (either that the animals never experience significant pain or distress, or that they always do). The point of having objective standards is to establish the facts, not to provide extra argumentative juice to the view you've latched onto for other reasons.

However, if there are lingering remnants of presumptive mistrust, then those arguing against research with animals may also distrust the "objective standards" the scientists offer to measure pain and distress and the efficacy of analgesia and anesthesia, viewing these as tests rigged to favor the scientists' agenda of continuing animal research. A pitfall of this move, though, is it takes these standards off the table in cases where animal responses seem really to indicate distress -- and if researchers aren't attending to animal distress, how can they alleviate it? (Also, staking out a claim that scientists are not to be trusted means that you cannot call on scientific literature to support you claims -- whether about animals' capacity for pain and distress or about animals' capacity for self-awareness.)

In this territory, we encounter another extreme position, that any animal interaction with humans counts as "torture", thus making all animal research animal torture regardless of the lengths to which individual researchers and IACUCs might go in trying to minimize animal distress and discomfort.

I would argue that, in the absence of a fuller argument (with empirical support drawn from observations of actual animals), such a definition strains the meaning of "torture" as commonly used -- and seems to smuggle in the emotional weight of the word as usually understood, substituting that emotional weight for a coherent argument. I think it's also a legitimate question whether labeling all human-animal interactions (or even all human-animal interactions in laboratory settings) as "torture" makes sense when the animals' experiences are compared to what they would typically experience in the wild.

And ultimately, I think the camp that asserts animal research is torture, even in the face of observations of the animals that reveal none of the responses consistent with pain or distress, is not really arguing about the facts of what the specific animals in question are experiencing. Rather, I think this camp's claim has to do with a philosophical commitment about the proper relation between humans and non-human animals.

A huge impediment to dialogue about the use of animals in research is an imprecision in laying out the philosophical commitments driving the people in the dialogue. That's the issue we'll take up in part 5 of the series.