A recent journal commentary entitled “Are e-Cigarettes Tobacco Products?” has caused a flare-up in the simmering bickering on the point. Such disputes are usually anchored, on one side or both, on the incorrect notion that there is some existentially correct rule for defining this (or any) category. The tone is often reminiscent of a child who just learned that tomatoes are a fruit, in terms of plant anatomy, arguing with his classmates who insist they are vegetables.
Everyone in that story exhibits a smug surety thanks to their failure to understand there are different definitions of the category “fruit.” Someone who just says “a tomato is not a fruit” is wrong. This is not because the statement is false but because it is meaningless without further clarification. Someone who says “fruits are sweet tasting parts of a plant, used more in desserts than savory dishes, so tomatoes are not a fruit” is right. A response of “that is not what ‘fruit’ means” is wrong because the category has different definitions and context matters. By contrast, a response of “this is a botany class, not culinary school, so that is not what ‘fruit’ means” is unassailable. Whether something is included in a category depends on what definition of that category is being used, and if that is not stated (and the context does not somehow make it clear), then it is meaningless to make assertions about it.
Categorizations like “tobacco product” serve several purposes. One is deciding which laws cover a particular product, company, behavior, or whatever. In most jurisdictions, there are a host of laws in place that apply to anything that is designated a tobacco product. Thus, the decision about whether to include vapor products in that category has major policy implications. But note the wording there, “decision to include.” There is no “right” answer; this is purely a policy decision. People with different political positions will feel that one choice or the other is better. It is certainly a good idea to fight against the expansion of abusive punitive regulations to vapor products (or any other products), and thus fight their inclusion in the legal definition. Just do not mistake the practical policy fight as having anything to do with some existential “tobacco-ness” of vapor products.
Part of the confusion is that this political debate often follows the common script of disguising a policy decision as a process decision. Tobacco controllers say “all tobacco products should be treated the same, so we should expand these tobacco laws to cover e-cigarettes” a statement about process. But the argument is really about whether a host of restrictions will be applied to vapor products. The response of “these are not tobacco products” — as opposed to “if you define the category in this way, it will impose those restrictions on vapor products and the bad results will be…” — falls into the trap of endorsing the position that the restrictions are a mere process rationalization. It effectively says “if you do not agree with my view, based on my personally-preferred definition of the term, that these are not really tobacco products, then you should go ahead and impose the restrictions on them.”
The more common use of a categorization is as a shorthand. For conversational usage a rough shared understanding is enough and the exact edges seldom matter. We all agree that “tobacco products” includes cigarettes and some similar items, and excludes coffee and the Golden Gate Bridge, and that is usually good enough.
However, sometimes a category is the basis for normative claims (“no one should ever use any tobacco product”) or scientific claims are made about the category (“there is no safe tobacco product”). When the category definition is doing any such work, it needs to be clearly defined or the claim is meaningless. No widely shared definition exists so it must be explicitly defined by whoever is making the claim. This choice of definition cannot be said to be wrong, so long as it is not silly, because it is being specified for a particular purpose and is not a claim that this is the only possible meaning of the term. However, since this basically never happens, normative or scientific claims about the category are usually nothing more than meaningless noise.
Compare the case of the category “planets.” Whether or not the category is broad enough to include Pluto is a decision. It does not have an inherently right or wrong answer, regardless of which way the official astronomy rulemaking body has most recently decided. Most conversations that refer to planets do not need to clarify the status of Pluto, let alone an exact definition. However, if a quiz question says “how many planets are in our solar system,” the answer is undefined unless it is stated, “according the definition…”
Some authors use “tobacco product” to refer to any consumer product that delivers nicotine (sometimes arbitrarily excluding products that are officially medicines, and sometimes not). Some use it, in effect, to refer to any product made by the major cigarette companies (even if the particular item is made by someone else). Others seem to think it should mean anything with bits of tobacco leaf, though they seldom make this clear. But how much of the plant needs to be there (some parts are always removed) and how big do the bits have to be? Do you have to be able to see them, or does a bit of powdered leaf invisibly contained in confectionary count? These are never clear. The WHO and FCTC, and tobacco controllers who take their cues from them, use “tobacco product” to mean “anything we wish to include in our portfolio of control,” which includes such things as South Asian dip products that contain little or no tobacco.
Personally, I tend to use it to mean “cigarettes and any other product that occupies the same drug-delivery and behavioral niche as cigarettes.” A focus on usage is far more useful for most purposes than making it about the object itself, since the interaction between people and the objects is almost always what we are interested in. The common use of “tobacco products” or just “tobacco” to really mean cigarettes (e.g., claims about the harms from “tobacco product use” are usually really only about smoking) offers a strong endorsement for anchoring the definition on cigarettes. But again — and this is the key to avoid getting caught up in silly arguments — in the rare cases where the terms is doing any work in my writing, I make sure to clarify how it is being used or avoid it altogether and just specify what products I am talking about.
One of the above candidate definitions excludes e-cigarettes, while most include them. A couple of the definitions include Zonnic (nicotine gum made by RJ Reynolds) but not Nicorette, while other definitions include both. The definition I prefer includes both when they are being used for their primary use (to deliver nicotine for people who want to keep consuming nicotine) but not when they are being used for what the manufacturers pretend is their primary use (temporary consumption as a smoking cessation method). Notice also that the definition I prefer excludes premium cigars and shisha to the extent that someone takes a few puffs on rare social occasions; those products are obviously made of tobacco leaf and deliver nicotine, but for most any practical matter under discussion (sensible policy, health and other impacts, habituation, and any but the most extremist “moral” concerns) this behavior is unrelated to cigarette smoking.
When a definition of a category is used to make claims about everything in the category, it should be subject to a few thought exercises. What if a science-fiction molecular assembly machine constructed a product that was physically identical to a Marlboro, but that did not use any tobacco leaf as a feedstock? Some of the proposed definitions of “tobacco product” would not include it. Similarly, what if we think beyond the existing heat-not-burn devices and imagine a device that takes bits of processed tobacco leaf and extracts and delivers only the nicotine? It seems difficult to imagine any possible normative or scientific claims that would apply to an actual Marlboro but not a replicator-Marlboro, nor to the hypothetical nicotine-only HnB product but not to a e-cigarette.
These thought experiments suggests that only a behavioral or functional definition can be useful for most scientific or normative purposes, though it need not be the specific version I prefer. Similarly the self-centered tobacco control definition, “anything we would like to control,” is functional and pragmatic, whatever else is wrong with it. It covers hypothetical new products and allows for blanket normative pronouncements, though they are effectively circular. However, it is important to realize that no legitimate scientific claims can be made about a category that is defined based on someone’s moralizing whim, whether that whim puts vapor products inside the category or out. Whenever tobacco controllers claim “tobacco products are X,” where X is any health effect or other material conclusion, they are talking nonsense. It is literally impossible that any such blanket statement could be true for their broad and open-ended political definition of the category.
One candidate basis for drawing the line that clearly does not work, though it frequently muddles the discussions around vapor products, is arguing that e-cigarettes should not be included because they do not have the health effects of cigarettes. Some products that are probably lower-risk than vapor products — American- and Swedish-style smokeless tobacco — are included in seemingly every candidate definition. If those products share the category with cigarettes, then obviously health effects is not part of the category definition.
Category definitions can be useful for a particular purpose. Different definitions can serve different purposes. Sometimes there is widespread agreement about a single definition among those who might wish to invoke the category, but not for the present question (or questions about “addiction” or any number of other concepts in this space). Indeed, about the only way to be bright-line wrong here is to insist that that vapor products definitely are, or definitely are not, tobacco products without stating you are choosing to use a particular candidate definition.