Thanks to Brian Tomasik and an anonymous reviewer for their comments. It was through reading Brian's writings that the topic was first brought to my attention.

Epistemic status: Establishing knowledge about other minds requires both science and philosophy. The scientific details should mostly be correct. The philosophical component uses plausible ideas – but it’s necessarily intuition-dependent whether we decide to accept or reject this kind of philosophical reasoning. Note that this subjectivity isn’t specific to thinking about microorganisms, but is rather general to speculating about the experiences of any mind other than one’s own.

Summary

Massive numbers of microorganisms, and their counteractive responses to various threats – in some respects similar to other organisms – suggest a possibility of immense suffering and an ongoing moral catastrophe. Despite this, few have seriously investigated the possibility of microorganism suffering. This post looks at the topic through both scientific and philosophical lenses. Before evaluating scientific findings, I generate, and highlight the importance of generating, criteria for such evaluation.

To demonstrate why microorganism suffering seems possible[1], I show that starting from ‘common sense’ ideas of what evidence of mind and suffering means, physically speaking, we can reach (to use Brian Tomasik’s words) ‘cosmopolitan views’ of evidence of mind and suffering that include systems that are more different. And it seems if we already accept ‘common sense’ views, we should consider ‘cosmopolitan views’ somewhat plausible. This is especially so if we have uncertainties with our subjective judgment – we might be compelled to assign at least some credence to ‘cosmopolitan views’, given some acceptance of ‘common sense’ positions. Additionally, uncertainty might suggest that we cannot be so confident in only ‘common sense’ positions as to rule out similar ‘cosmopolitan views’. That said, whether we accept or reject these arguments also seems intuition-dependent.

I note some examples of microbial behaviors and mechanisms that fit the cosmopolitan criteria. If my arguments are successful, forcing us to place some credence in cosmopolitan views, then there may be significant amounts of microorganism suffering in expectation. We may have reason to try and understand more about how we can reduce such suffering.

A video showing two paramecia eaten by amoeba:

Introduction

At any given moment, around 1030 (one thousand billion billion billion) microorganisms exist on Earth.[2] Many have very short lives, resulting in massive numbers of deaths. Rough calculations suggest 1027 to 1029 deaths per hour on Earth.[3] Microorganisms display aversive reactions, escape responses, and/or physiological changes against various fitness-threatening phenomena: harmful chemicals, extreme temperatures, starvation, sun damage, mechanical damage, and predators and viruses.

It’s no exaggeration to say that if microorganisms can suffer, such suffering is immense in quantity. Even if microorganism suffering is unlikely, we might still argue that it is extremely large in expectation.[4]

This possibility of large-scale suffering may warrant further exploration. Yet, aside from a few exceptions, the topic has been severely neglected. My first impression is that reducing microorganism suffering is a cause that some should potentially prioritize, from the perspectives of many non-speciesist, welfarist views. I won’t be investigating cause prioritization in detail here. Instead, in this piece, I only aim to explain why I think microorganism suffering is at least a possibility. By ‘possibility’ I mean not astronomically lower in probability compared to the existence of suffering in more recognizable systems (see footnote for details).[1] Ideally, this work would motivate further discussion and investigation.

I first provide definitions of “microorganisms” and mentally experienced “suffering”, and explain the ever-present challenge posed by the ‘problem of other minds’. Next, I explain why it’s important to both figure out criteria of what we mean by evidence of suffering, and look at the available scientific evidence based on such criteria. I proceed to construct criteria for evidence of suffering and conclude that if we are to give some credence to more ‘common sense’ positions we seem to be required to also give some credence to more ‘cosmopolitan views’ of suffering. I also suggest that we cannot be too confident in the idea that only ‘common sense’ perspectives are true as to rule out extended cosmopolitan views. Finally, I mention findings from scientific experiments and observations which meet such cosmopolitan criteria, suggesting that microorganism suffering is a possibility that we cannot ignore. (In Appendix: Factors affecting how we attribute suffering, I also describe various biological, social, and psychological factors that affect how we attribute suffering to physical systems.)

Paramecium (prey) trapped inside Heliozoa (predator).  ‘A sol mit Paramecium’. Image by M. Linnenbach. (from https://commons.wikimedia.org/wiki/File:A_sol_mit_Paramecium.jpg) This file is licensed under the Creative Commons Attribution-Share Alike 3.0 Unported license.

Microorganisms discussed

By ‘microorganisms’ I refer to single-celled prokaryotes such as bacteria and archaea, and eukaryotes such as protists, unicellular algae, and yeasts. I don’t include nervous system-possessing microscopic animals[5] such as nematodes. Even though viruses are sometimes considered microorganisms, I avoid including them in this analysis as they’re too different in structure and function. That said, the possibility that these other types of biological – or more generally, physical – systems suffer also seems worth exploring.

Suffering in mental terms and “The Problem of Other Minds”

We can describe suffering in both physical and mental terms.[6] For example, if I accidentally step on something sharp my body (including my brain) would physically react by activating nociceptors, firing neurons, and attempting to withdraw from the source of damage. While that’s happening I also have negative mental experiences.

For this post, I follow Mayerfeld (1999) and define suffering in mental terms as an overall bad feeling. This is a useful starting point but it’s not enough. A problem with discussing other minds is that we simply aren’t them. Knowing what suffering is in mental terms only demonstrates our own ability to suffer.

One author puts “The Problem of Other Minds” succinctly:

Start with yourself. You experience pleasure and pain. You can be as confident of this fact as you can be of any fact. Why? You have direct introspective access to at least some of your phenomenal states. But there is an asymmetry between you and everything else. You cannot know by direct introspection that someone else is phenomenally conscious. If you are justified in believing that other entities experience pains and pleasures, it must be by some different epistemic strategy.

We know by direct introspection that we can suffer. But we cannot know by direct introspection that others can suffer. We must use “different epistemic strategies” to try to find out whether others can suffer.

So what are these different epistemic strategies? Besides our own mental world, we have access to a shared physical world – where we locate other entities. To find out whether others can suffer, it seems we must refer to physical reality. To attribute suffering to others, physical evidence of mentally experienced suffering are things we must refer to.[7]

The importance of both finding and defining

Evidence of (being capable of) suffering allows us to claim sensibly that a system can suffer. This requires two things. We need criteria of what counts as evidence of suffering (defining), as well as details of the system that can be interpreted as evidence according to such criteria (finding).[8]

Discussions about whether a given physical system can suffer often focus on finding something in that system that corresponds to “evidence of suffering” rather than defining in physical terms what “evidence of suffering” would look like in the first place. For instance, when considering claims of animal suffering, discussions are often about whether they match humans in neurological or behavioral terms. People are often quite confident in their ideas and definitions of what counts as evidence of suffering. A common, but possibly unfortunate, attitude seems to be that all that’s left to focus on is finding out those selected neurological or behavioral similarities in animals and other systems.

Yet both finding and defining are essential. On the one hand, if we don’t find anything in the world and there’s nothing to refer to, then we lack evidence of suffering. And on the other, if we lack criteria, even implicitly, there’s no sensible notion of what “evidence of suffering” even means.[9] Focusing on one to the neglect of the other could be a huge mistake.

Figuring out criteria/definitions

Summary dialogue

This section is a bit tricky to understand. The following hypothetical dialogue might help summarize the key ideas less formally and allow you to skip ahead to the scientific findings. I don’t include nematodes as “microorganisms” in the definition used in this post but they provide a starting point for the debate between Alice and Bob:

 

Alice: “Do you think nematodes can suffer?”

Bob: “I know nematodes can’t suffer.”

Alice: “How do you know that?”

Bob: “Well, I read this article about how scientists failed to find evidence of suffering in nematodes.”

Alice: “What did the scientists mean by ‘evidence of suffering’?”

Bob: “Wait, let me look it up… (Bob searches the internet) Okay, it says here that by ‘evidence of suffering’ they meant XYZ”

Alice: “Right, but are those the right criteria?”

Bob: “Of course – I mean they probably figured it out scientifically.”

Alice: “But the scientists can only know suffering from a first-person perspective. How did they figure out how to describe it in physical, third-person terms?”

Bob: “Alright, okay. Maybe it is partly philosophical. Well, it seems reasonable to assume their brains/nervous systems have something to do with their suffering. Information from experiments manipulating the other (similar) systems of other humans and animals and seeing what’s required for suffering helps them find out how to describe what’s necessary for suffering.”

Alice: “Proceeding to assume what’s necessary for human/their suffering is the same as what’s necessary for the suffering of other non-humans/others is a big leap.”

Bob: “What do you mean?”

Alice: “I mean even if you establish that ‘if an important brain part then suffering’ (I’m oversimplifying the neurological details here) it doesn’t mean that ‘if no important brain part then no suffering’. That’s a fallacy (denying the antecedent).”

Bob: “Hmm, I guess so. But that doesn’t mean we can say we are as confident in the suffering for ‘not-brains’ as we are in brains.”

Alice: “Of course, I agree. It seems we should have higher credence that brains are relevant to suffering than in the relevance of ‘not-brains’”

Bob: “But you still seem to think some ‘not-brain’ things could be relevant?”

Alice: “Yes, I do. I think we often overemphasize the very specific details of the human brain and nervous system in discussions of the possibility of suffering at the most basic level (basic as in lacking the ‘fancier’ aspects of human sentience surrounding it, e.g. ‘simple’ pain). It seems plausible to me that many other forms of information processing are similar enough to the general picture to be sufficient for suffering at a basic level. We can’t rule that out at the very least and I think we might be biased by looking at things from a human perspective.”

Bob: “But we can’t be that lenient! We should just stick with common sense views instead.”

Alice: “Well, ultimately we can’t justify even the ‘common sense’ views. Considering other people with similar brains as conscious and capable of suffering is a subjective choice. Deciding that other people are similar enough is subjective. A human who’s super strict about what’s considered sufficiently similar could deny that other human brains are similar enough to their own to suffer. They could be a ‘suffering-solipsist’ (who only believes in the possibility of their own suffering). We could be ‘suffering-solipsists’.”

Bob: “That seems wrong.”

Alice: “Ah, but maybe you’re being too lenient!

Bob: “...”

Alice: “On the other hand, if we’re being more cosmopolitan/lenient, we might also include other biological systems such as single-celled microorganisms.”

Bob: “Alright, but how do we tell who’s right? How similar is sufficiently similar?”

Alice: “Good question! Obviously, it’s hard to tell. And an important thing to remember is that it seems there’s no escaping subjectivity. I suppose we might find it intuitive that it’s more likely for a dog to be similar enough to suffer than an insect, more likely for an insect to be similar enough to suffer than a nematode, more likely for a nematode than it is for a bacterium, and more likely for a bacterium than it is for a simple virion. If we think we can justify or have faith that some of those systems (such as other humans and animals) do suffer, we might also give some credence to the suffering of slightly different systems, and adopt a 'cosmopolitan view'. We humans also potentially have a few biases that affect our thinking about what can suffer. We’re uncertain and one might also argue that with such uncertainty, given that we’re confident about our current ‘common sense’ ideas, it seems it’s difficult to be overly confident that only ‘common sense’ views are true to an extent that effectively rules out the alternative options with slightly different thresholds for acceptable similarity. In the case of microorganisms, approaching their scientific details with these cosmopolitan views (that we assign at least some credence to), we might conclude that microorganism suffering is possible.”

 

You can skip to the science or read the philosophical justification for 'cosmopolitan views' in more detail:

 

Unsatisfying categories

How do we figure out what counts as evidence of suffering? Or at least, how do we establish preliminary ideas of what that might be? I’ll explore one common approach.

Philosophers have argued for the attribution of pain (and mind) through:

  1. Argument by analogy (“I react to damaging stimuli and feel pain. Others react to damaging stimuli, and analogously they also feel pain”).
  2. Inference to the best explanation (“The best explanation for others, who are similar to me, reacting aversively to damaging stimuli is that they feel pain, similar to how I do”).

Here, we assume that one or both of these forms of argument work as it doesn’t seem that that’s always the case.

Notice that such arguments require us to identify the ‘others’ to which we apply the arguments. Intuitively, we might require the ‘others’ to be physically similar to us in certain ways, so as to claim that these ‘others’ are mentally similar to us in certain ways. 

What exactly is it that’s physically alike between ‘others’ and ourselves that allows us to claim they can feel pain like we do? As humans, pain seems to have to do with “relevant stuff specific to our nervous system”. Modern neuroscience and a long history of medical observations show us that something about it is relevant. E.g., blocking mechanisms of the nervous system, anesthetics and analgesics reduce or eliminate our experience of pain.

For us, as the biological machines defined by “relevant stuff”, “relevant stuff specific to our nervous system” are likely important for our experience of suffering.

It’s tempting to conclude things here and just say, for example, systems with “relevant stuff specific to our nervous system” have evidence of suffering, but that systems without them do not – and might even show evidence of a lack of suffering. This may be a case of denying the antecedent. We might instead suspect that things aren’t that clear cut. “Relevant stuff specific to our nervous system” is an invented category to which we assign actual things, that is, particular individual relevant things. For instance, we’re the ones subjectively grouping together the brains of different humans as “human brains”, and the nervous systems of different animals as “nervous systems”.

The categories are human constructs. It isn’t the case that there’s a special label attached to our brain that says “human brain” – fundamentally distinguishing it from everything else. Individual humans are “clustered together” by shared traits, and we assign the label “human” when we feel it’s appropriate to. Another example: we observe that “humans” have an important “organ” in their “heads”, and group those individual objects under the label “human brain”.

The categories are fictions that are sometimes useful. In a scientific context, it’s often helpful to be able to speak of “brains” and “nervous systems” and how they work by themselves and in relation to other parts of the world. What matters is whether such fictions are useful in the context of revealing suffering, wherever it may be.

But here lies the problem. Boundaries of science-suitable categories don’t seem entirely suitable for our quest to identify the categories of conscious/sentient/suffering-capable systems. Why should they be? In fact, the categories (and the boundaries that define them) are most likely non-identical.

Nothing says that “invented categories that make it convenient for humans to discuss science will necessarily reveal all that there is to know about the mental experience of suffering”.

A separate category for “systems capable of suffering” needs to be constructed – with its elements fitting acceptable criteria. Although we still might want to keep “nervous systems” in mind since our suffering seems centered on it. It’s just that – the boundaries probably aren’t identical.

We need to define new boundaries for our new category of “systems capable of suffering”.

But it turns out, it might just be that: whether we could recognize others to be as similar to ourselves in our suffering-relevant bits and group them together with us as “systems capable of suffering” depends on how strict or lax we subjectively are.

To illustrate, we might intuitively and roughly claim, in the order from most similar to least[10], “relevant stuff specific to our nervous system” are followed by:

  1. “Relevant stuff” specific to the nervous systems of other humans.
  2. Of other non-human mammals.
  3. Of non-mammalian animals
  4. Other forms of cell-to-cell signaling such as nervous system precursors
  5. Signaling within a single cell such as with microorganisms.

Which systems do we include/exclude when constructing the category “(biological) systems capable of suffering”? How similar is similar enough? The problem of other minds comes up again. Not being other minds prevents us from directly knowing. Lacking that objectivity, deciding on a decision boundary is a subjective process. It seems that we must subjectively decide using our intuitions – intuitions that might be quite independent of the reality of what can actually suffer. It might be that there is actually no good way to tell. (See the relevant section “Appendix: Factors affecting how we attribute suffering”)

Cosmopolitan views

There might be no clear way to establish a boundary between (biological) sufferers and (biological) non-sufferers. Different people will have different ideas on what counts as sufficiently similar to count as being capable of suffering – and it’s hard, if not impossible, to tell who’s right, if anybody is.

Some might still think that ‘common sense’ claims about who/what can suffer (E.g., “We know that some other humans can suffer”) must be self-evident/provable despite this uncertainty. I’d disagree – but we do seem to in effect, at least, accept the ‘common sense’ standpoints. Not being others means we can’t confirm that they feel pain, and our method of arguing that they do feel pain is riddled with the use of intuitions. But despite the philosophy, we seem to have faith that they do (e.g., faith that other humans do suffer), and act in the world accordingly.

The following argumentation works best assuming that we place some weight to common sense positions (positions such as “We know that some other humans can suffer”).

Noting the unclear boundaries, and accepting common sense claims to some extent, we might expand into ‘cosmopolitan views’ – additionally including further biological (or more generally, physical) systems.

For instance, here’s Brian Tomasik expressing a more ‘cosmopolitan view’ that I tend to share:

The basic idea is to point out that, at a high level, we are all just stimulus-response machines—receiving inputs from the environment and producing changes in physiology and behavior in response. What's different among [e.g. different living things] is the degree of complexity involved. Humans engage in far more detailed processing of input stimuli into more complex mental representations, and have more sophisticated internal self-reflection and self-stimulation systems. But as far as I can see, there aren't discontinuities here. There's not a single cognitive architecture that's required before the "lights of consciousness" turn on, where before all was darkness. Rather, there's a vast space of possible mind designs, some with greater detail and self-reflectiveness than others.

This ‘cosmopolitan view’ seems to potentially follow from the more ‘common sense’ ones. We’ve only dialed down our threshold[11] for acceptable similarity. We follow the trend of “thinking that other humans are sentient and can suffer, then thinking that other large and similar non-human animals are sentient and can suffer, to then thinking that small animals such insects are sentient and can suffer”…and so on. This next stage only relaxes the criteria and allows somewhat simpler and different information processing to be included as evidence. Indeed, initially, we focus on a form of cell-to-cell signaling (what nervous systems do), and now we also include within-cell signaling. This in turn allows for biological systems with slightly more difference to be included in the category of “(biological) systems capable of suffering”.

My intuitions make me lean towards thinking that such a view makes a lot of sense. I personally would think that it’s likely acceptable criteria. You might not share the same degree of confidence of course. But note that we’ve reached this view by reducing a threshold that we were already reducing, and so it intuitively seems that such a view is comparatively plausible. 

(Alternatively, someone with a very chauvinistic requirement of similarity might hold that nothing else is similar enough to themselves. They might claim that nobody else can suffer and adopt a position of ‘suffering-solipsism’.)

Of course, a potential issue with somebody expanding their intuitions further based on intuitions is that they’re doing it based on intuitions, which may be unreliable. This perhaps is the biggest challenge to this view. If we thoroughly reject such intuitions, in favor of other intuitions, then we find ourselves dismissing the possibility of microorganism suffering.

Intuitions vary from person to person. Again, there might be no good way to tell who’s right, and it’s unclear what the correct subjective approach is, so to speak. Nonetheless, as explained, this ‘cosmopolitan view’ might be one somewhat relatively (intuitively) convincing alternative – potentially adjacent to common sense views.[12]

One might also argue that our uncertainty might make us assign some credence to cosmopolitan views. With a good deal of uncertainty from recognizing the subjectivity of determining what counts as sufficiently similar, coupled with the risk of various possible biases (Again, see the section “Appendix: Factors affecting how we attribute suffering”), it seems that if we decide to assign some credence to more ‘common sense’ views, then we should also assign some credence to closely related cosmopolitan views. Similarly, it seems that we can’t be too confident that only ‘common sense’ positions are true, to an extent that rules out alternative views such as more cosmopolitan views.[13]

Overall, we might assign at least some credence, if not more, to such cosmopolitan views that focus on the high-level perspective of conscious and suffering-capable systems as stimulus-response systems that differ in complexity. We might accept that there’s some evidence that suffering exists whenever a system “simply” reacts to avoid actual or potential damage to itself, with better evidence when there are fancier “complex” things going on, e.g. learning. 

To evaluate scientific findings using this perspective and conclude there's a possibility[1] of microorganism suffering, we don't necessarily require full confidence in this view – it's sufficient to think it could be true with some likelihood. (Personally, I think it's ~20-30% likely but this could be on the higher side for most people.)

Counterarguments

One counterargument might claim that neuroscience does tell us that certain things of the nervous system are necessary for consciousness and suffering and what is irrelevant, e.g. allowing us to claim that some responses are ‘mere reflexes’.

Studies that attempt to find a ‘minimal set’ of neural correlates often choose some outputs to focus on and count those as indicators of pain – often more ‘notable’ processes, e.g. memory, verbal report (3rd person), or introspection (which is 1st person – though always reported through 3rd person processes). The parts necessary to activate those outputs might be deemed the ‘minimal set’ of neural correlates of pain, and those not needed might be deemed irrelevant to pain.

It’s a stretch to say we can identify what’s irrelevant to pain. The problem here is that what outputs to highlight is a subjective decision influenced by a ‘reportability bias’. We’re choosing these “more famous” outputs from our point of view as the conscious systems tightly linked to specific functions producing those outputs, e.g. memory, verbal report, and introspection. So of course we consider those outputs to be more reliable reports of pain! But that doesn’t mean we can conclude other aspects of the nervous system are irrelevant to pain. Sure, they may be irrelevant to our pain, as conscious systems tightly linked to memory and verbal report, but it’s unclear whether they are irrelevant to all pain.

For instance, we haven’t ruled out their relevance to what we might experience if we hypothetically became ‘simpler’ and lacked a range of our abilities.[14] This reportability bias limits what the search for neural correlates of pain can label as being irrelevant to pain. By being sufficiently similar to relevant things, these supposedly irrelevant things might still turn out suffering-relevant, e.g. they aren’t ‘mere reflexes’ to ‘less complex’ minds.

Another counterargument is that there are strong examples of parts of the nervous system that show what’s necessary for pain. For example, one could concretely point to the SCN9A gene that encodes Nav1.7 sodium channels – which enable generation of action potential, note that mutations of the gene that cause a loss of function in the channel protein results in reported insensitivity to pain (in the usual sense of a report), and conclude that a functional version of the gene/channel is necessary for pain.

A potential problem is that the argument is context-dependent. A functional version of the gene/channel does indeed seem necessary for pain in the context of the “default” system of a human being. Yet, this might not be the only way pain could come about. It seems plausible that if humans were hypothetically “modified” to generate action potential anyway, or to connect the rest of the nervous system to nociceptor replacements, or to possess a somewhat redesigned nervous system, all different ways of bypassing the need for functioning Nav1.7 channels, they could still report on pain (in the usual sense) – which we may take as strong evidence of pain. The overall function might still be replicated in other ways. We could perhaps argue similarly for other nervous system parts, such as pain-mediating circuits in the brain (although I understand less of the details). We might also think that replication of function could vary to different extents of accuracy while being similar sufficiently for there to be a possibility of suffering. It’s not clear that some things are necessarily necessary.

What about differences that might be more fundamental? Sometimes the ability for self-reflection/higher-order processing is brought up as an essential requirement for conscious awareness and hence suffering. There are some responses to this view. First, we might point out that it might feel like self-reflection is required for consciousness mostly “because we never see consciousness apart from higher-order reflection”. We might consider “higher-order thought [as a] "measurement device" that's always present when observing” our “low-order experiences”. Even if we accept the requirement for higher-order reflection, it may be the case that “processes reflecting on processes” is actually ubiquitous.[15]

One good counterargument is that we should also recognize what’s considered “not in pain”. A reported insensitivity to pain (in the usual sense; and note that this is distinct from a lack of a report) is good evidence that the system, considered as a whole and ignoring subsystems, is in a state of not being in pain. Using this information we may have some idea of what “not in pain” looks like. There might be many other systems similar to the system that's not in pain, and we could argue from the similarity that the other similar systems don’t experience pain. I think this argument is strong. It does limit our interpretations of what suffers to a large extent. However, this has to be balanced with the above-mentioned arguments that suggest that systems similar to systems we know to be capable of pain, also can experience pain. It’s again not clear that we can firmly define a boundary that excludes microorganisms such that we can confidently say there’s no possibility of microorganism suffering.

Scientific findings

Science helps us find things that might fit the criteria that we consider somewhat plausible. What has science found?[16]

There’s a lot of literature on various damage-reactive aspects of microorganisms. It’ll take far too long to summarize everything scientists know – and I doubt that I’ve personally explored everything available. I’ll only cover some phenomena that might be of interest at a superficial level.

The following descriptions will be simplified. Language necessarily limits the ability to capture molecular, biochemical, and ecological complexity and sophistication understood through scientific experiments and instruments. Also, I’m just posting this on a forum.

Bacteria chemical-reactive movement (chemotaxis)

Flagellated bacteria use ‘tumbles’ to change swimming direction. This ‘tumble’ – the basis of escaping potential harm and associated chemical signals – is caused by the interaction of proteins at the bacterial motor on one end of the bacterium, which is induced by pathways of other protein interactions, in turn, triggered by detection of chemical repellents or lack of detection of attractants by a sensory array of ~10,000 receptors (a ‘nanobrain’) at the other end of the cell. A kind of short-term ‘memory mechanism’ also exists to detect continued presence of repellents or attractants, creating a biochemical negative feedback loop that takes note of the recent past and prevents over-tumbling or under-tumbling, presumably helping with adapting to more harsh or more welcoming environments.

In greater detail, chemical-reactive movement and other bacterial behaviors has been described by Lyon (2015):

Like all organisms, bacteria adapt to changes in their environment by modifying their metabolism and behavior. The traditional view is that this process depends on a limited number of highly canalized, inflexible mechanisms, which bear little resemblance to their behavioral counterparts in animals with nervous systems. However, evidence increasingly points to the idea that bacteria—particularly ecological ‘generalists’ as distinct from ‘specialists’—are highly social, flexible responders, responders (however limited their actual response styles) that rely on a sophisticated suite of sensory and information-processing mechanisms (Shapiro, 1998). Such mechanisms are characterized by non-linear responsivity, integration of multiple information channels, and (in the case of [chemotaxis] receptors, at least) habituation and adaptation phenomena. These structural and dynamical features are among the defining features of neuronal sensory processing in animals, including humans.

Keep in mind that a great deal of complexity will be obscured in the simplifications that follow. In each cell of E. coli, for example, which is not the smartest proteobacterium on the block (but no dummy, either), there are upward of 10,000 chemoreceptors per cell, each with multiple binding sites, whose output interacts with several flagella, on each of which are about 40 binding sites for CheY-P, the protein that modifies the direction of rotation. The behavioral output for a single cell is thus hugely complex, and much more refined (tumbles, twiddles, etc.) than normally described. When we simplify to get just two outputs, run and tumble, and then bundle behaviors to get population averages, a lot of the cognitive output is likely to be occluded. Consider an example with humans: if we assumed just two behaviors, walking and standing, and measured population averages using different stimuli, more would be hidden in the results than is revealed…

...the assessment of valence by the cell that underlies [chemotaxis] in the presence of both has long been known to be non-linear even in simple laboratory setups (Adler and Tso, 1974)...

…Long-term memory is often indistinguishable from non-associative learning, a type of learning in which “presentation of a particular stimulus alters the strength or probability of a response according to the strength and temporal spacing of the stimulus” (Rosenzweig et al., 1996, p. G-17). Non-associative learning includes sensitization, the amplification of a response following presentation of a stimulus, and habituation, the attenuation or extinction of a response to a stimulus upon repeated presentations (Shettleworth, 1998). Habituation and sensitization have both been demonstrated in bacterial [chemotaxis] (Koshland et al., 1982; Stock, 1999; Porter et al., 2011)—a discovery that “gave some neurophysiologists apoplexy, because they believed that a nervous system” was required (Taylor, 2004, p. 3761).

Cellular stress response

Severe stresses, such as extreme temperature, toxins and mechanical damage, go beyond usual cell maintenance capacity limits and damage a cell’s macromolecules[17], thereby disrupting the proper regulation of its processes.

To deal with this, stressed cells in ‘all domains of life’, bacterial, archaeal and eukaryotic, activate a ‘cellular stress response’ – a ‘systems level phenomenon’ involving many cell mechanisms – to restore ‘structural and functional homeostasis’ in the stressed cells. This involves large-scale rewiring genetic networks and networks of interacting proteins within the cell.

It’s interesting that ‘simple’ prokaryotes also have this response. The bacterial stress response[18] can result from lacking food, damage from other organisms, and abiotic sources. Referring to the bacterial stress response and related phenomenon, one biologist notes that: “[t]here’s a tendency to think of bacteria as like little robots with a fixed strategy that just do one thing but they’re extremely responsive, extremely sophisticated organisms that can detect their environment and respond in very very sort of what we would consider to be sensible ways such as fighting back when attacked or just like hunkering down when faced with something like sun damage.”

(It’s impossible to explain everything that goes on in cells in a short post. Also, scientists might never map out every mechanism that occurs in every organism. If you’re interested you can read reviews like this one.)

Predation, parasitism and viruses

Predators and viruses in the microbial world may provoke damage-avoiding behaviors and mechanisms.

Some examples[19]:

  • Bacteria activate defense mechanisms to respond to protist threats.
  • Some ciliates (single-celled eukaryote) display a rapid escape swimming behavior in response to microscopic predator animals..
  • Archaea and bacteria possess CRISPR-Cas systems, which can help defend against viruses and memorize some of their molecular signatures for better future defense.
  • Bacteria may release antibiotics that kill predator bacteria as well as enzymes to deactivate harmful antibiotics released by predator bacteria.
  • Bacteria engulfed by amoeba (single-celled eukaryote) utilize strategies to survive.
  • Some yeasts (single-celled fungus) engulfed by amoeba can escape.
  • Amoeba feeding strategies seem to take into account the possibility of ciliates escaping. This video shows a ciliate seemingly failing to escape from being eaten alive.
  • This video shows a paramecium using a trichocyst for defense after sensing a homalozoon predator.

For a visual understanding, these two videos[20] show instances of predation under a microscope. Some moments may show microanimals rather than microorganisms as defined.

Single-celled eukaryote capabilities

This is an indirect argument. While it’s hard to describe the many details of various eukaryotic response mechanisms, the surprising capabilities of single-celled eukaryotes might indicate more sophistication in general, and suggest underlying complexity.

On the complexity of protists, Moselio Schaechter writes in the foreword of the ‘Handbook of the Protists’:

Cellular complexity is carried to extremes by the protists. The ciliates and flagellates (mastigotes), for example, possess – to shamelessly use anthropomorphic terms – a mouth (some with fancy lips), a stomach, an anus, a bladder, multiple propulsion devices, and an armamentarium of weapons that allow them to feed on other microbes. And some dinoflagellates sport that most amazing of cell structures, the eye-like ocelloid. In multicellular organisms, such structures and their functions are assigned to specialized cells; in the protists, one cell does it all by itself. One wonders about the trade-off involved, in doing everything yourself versus employing a differentiated consortium to do the work. The very existence of protist complex body plans brings up the distinction between unicellularity and multicellularity, a blurry one at best, but one especially relevant to this field. Multicellularity has originated on multiple occasions within the protists, hence they are most likely to provide relevant answers to our conjectures.

The available scientific literature is vast. No doubt I’ve missed out on some important findings. In addition, scientists don’t entirely understand cellular mechanisms and abilities of microorganisms. The scientific conclusions listed do match the criteria in different ways. If we have at least some credence in the cosmopolitan criteria, then we do have some evidence for the possibility of microorganism suffering.

Concluding thoughts

“The fact is, I don’t even know that you’re conscious.”Christof Koch

Knowledge of other minds and their capacity to suffer will always rely on subjective understanding. Our strategies of developing suitable criteria for evidence of suffering are intuition-dependent and are susceptible to further intuition-dependent changes. Combined with scientific knowledge of the systems' structure and/or function, it might lead us to think that it’s likely, or that there is at least some chance, that these unfamiliar systems can suffer. Additionally, as with other organisms, microorganisms display a variety of aversive reactions and physiological changes in response to threats, implemented through molecular networks of surprising complexity.

Failing to rule out the possibility of suffering in microorganisms might imply a large amount of microorganism suffering in expectation and a host of ethical obligations that we find burdensome. But perhaps this is no good reason to ignore our newfound thoughts. Instead, this might be a reason to attempt to understand how to reduce such possible suffering.

Other resources

Brian Tomasik’s ‘Essays on Reducing Suffering’ have had a large influence on my views of consciousness, suffering and the possibility of microorganism suffering. Pieces that discuss the possibility of microorganism suffering include ‘Cognitive Abilities of Unicellular Organisms’ and ‘Bacteria, Plants, and Graded Sentience’.

Pamela Lyon’s 2015 review of bacteria behavior highlights parallels between bacteria function, mechanisms, evolution and ecology and the concept of cognition as understood by cognitive scientists.

Journey to the Microcosmos is a YouTube channel featuring narrated, high-resolution videos of microscopic life. Some content may be distressing to viewers more sensitive to possible microorganism suffering.

Appendix: Factors affecting how we attribute suffering

It might be helpful to first understand factors affecting how we humans attribute suffering and mind. Understanding these factors may reveal possible biases, help us understand that biological, social and psychological influences could play a strong role, and might suggest that we shouldn’t be too confident in our immediate assessments.

Human intuitions were shaped in part by our evolutionary past. Like abilities to attribute other mental states, the capacity to attribute “suffering” to other humans/human ancestors was probably helpful in social interactions with other humans/human ancestors. It hence made such an ability beneficial to pass on from an evolutionary perspective. To some extent, features not unique to humans and shared with other (macroscopic) animals, e.g. appearing hurt, seem to also trigger some of these intuitions too.[21] As a consequence, many of us believe and feel a sense of certainty[22] that other humans and some other animals can suffer.

However, evolution might not produce intuitions directed towards other kinds of systems, whether or not those systems can suffer. Many other systems are relatively less important to the evolutionary fitness of humans and their ancestors (e.g. ants). Some systems might more easily be understood by a “physical stance” (e.g. falling rocks) and don’t require applying an “intentional stance”. Moreover, many other systems (e.g. microorganisms) might not register at all to insensitive human (ancestor) sensory faculties.

It doesn’t seem like human intuitions were necessarily evolved to track the internal experiences of other systems – but even if we suppose brains did do so (which, again, seems doubtful), then evidence of suffering at the microscopic level, if it exists, is never detected, and cannot be used to tune human intuitions about suffering. Either way, human brains would not evolve to recognize (microscopic) evidence of microorganism suffering whether it exists or not. A lack of intuitive acknowledgement that microorganisms can suffer might not be evidence to think microorganism suffering doesn’t exist, because intuitions about such suffering wouldn’t evolve.

Societal factors might also partly explain our beliefs. For most of human history, humans were oblivious to the existence of microorganisms and their abilities. Microorganisms weren’t observed until the 1660s to the 1680s. It wasn’t until the 19th century that Europe more widely accepted germ theory. The field of modern molecular biology only took off in the 20th century. Our 21st century understanding of microorganisms is still far from complete and we don’t frequently observe the microscopic world nor actively discuss the capabilities of microorganisms and their underlying mechanisms. If we think that knowing more about microorganisms helps us better find out whether they can suffer, but knowing less keeps us uninformed about that possibility, then it seems that we’re not justified in being confident with our current assessments.

A more fine-grained perspective shows us that different individuals and communities have different approaches to the attribution of suffering, and more generally, of mind. Philosophers disagree about how minds work and neuroscientists feel the need to revise assumptions on how brains relate to the mind. Among altruists there is discussion of invertebrate sentience and suffering, which some find plausible and others find implausible. Human “theory of mind” also varies with neurological differences. It also seems plausible that individuals in social groups may reinforce each others’ beliefs about mind and the capacity to suffer. In fact, in current times, it’s definitely socially embarrassing to publicly hold the view that microorganisms can suffer. We should expect social pressures to work against the idea, whether or not it’s true.

Furthermore, we might suspect that Western civilization’s influential, dualistic notion of “immaterial souls” has heavily shaped thinking about the mind. Concepts of “souls” and of “minds” are similar (and sometimes confused). Both relate to the concept of personal identity, are said not to be directly observable, and are sometimes used to explain behavioral complexity. One human, rather than non-human, life counts for one soul. This approach isn’t universal. Many other cultures have traditionally had different views.[23] Given Western influence on the world and the beliefs of scientific and philosophical communities, we might expect the concept of souls to have had a large impact on our beliefs on minds. On the other hand, there is a common human tendency of projecting our own minds onto the world, as in animism and religions, and that might also influence our thinking.

Biological, social, and psychological influences don’t necessarily guide us to the truth of the matter. This may suggest that we should avoid being too confident about dismissing such a possibility.

Appendix: Maybe the suffering isn’t intense?

One might occasionally encounter the view that even if microorganisms can suffer, we should expect microorganisms to suffer at a very low intensity. While speculating on the intensity of microorganism suffering is beyond the scope of this post, I’ll say a few things here.

The fact that there’s more going on in a brain than in a microbe does suggest some arguments in favor of there being more going on for a brain than for a microbe – but there are also arguments that suggest a more equal weighting of biological entities.

For instance, In humans and other animals, extreme suffering plays a crucial role in forcing organisms to be sufficiently “motivated” to respond to life-threatening emergencies. Mild suffering can often be ignored. There seem to be many examples where microbes “can’t ignore” stimuli that are damaging/signal damage. If this aspect of extreme suffering is applicable to microorganisms, it may be indicative of its presence. A related point is that perhaps relative signal strengths matter more to organisms than absolute signal strengths.

There’s an extensive report on the intensity of valenced experience across animal species. Not every argument concerning animals can be applied to the microorganism case – but it seems possible to do that for quite a few points. The report concludes that it’s unclear how the intensity range of non-mammalian animals compare to that of mammals (which includes humans), i.e., it could be narrower (less intense suffering) or it could be wider (more intense suffering). This uncertainty is partly because of “[increased] phylogenetic distance” (evolutionarily far apart). We might see microorganisms in the same way: quite evolutionarily distinct, suggesting that it’s difficult to say how the intensity of their suffering compares to ours (might be more or it might be less).

The intensity of suffering experienced matters may be especially important to lexical views of reducing suffering where the reduction of lexically worse extreme suffering is prioritized over the reduction of mild suffering. If microorganisms suffer mildly then it might be that those accepting such lexical views wouldn’t think microorganism suffering is dominating. However, as explained, the matter of the intensity of possible microorganism suffering is very unclear, so it’s also unclear what lexical views would imply.

Besides intensity, calculations of suffering associated with microorganisms, humans and non-human animals could be affected by our choice of approach to counting conscious subsystems. Similar to the other philosophical questions explored in this piece, it’s unclear which approach is appropriate and our choice seems dependent on intuitions. It’s also unclear which algorithms to implement the counting approaches. And whether counting even makes sense is questionable, e.g., what do we do if physics turns out to be not discrete?

  1. ^

    By ‘possibility’ I specifically mean that I think our credence of microorganism suffering shouldn’t be astronomically lower than our credence of human and non-human mammal suffering. It’s hard to put a number on this, but I think by that I mean at least not lower by more than ~4 orders of magnitude (not lower than ~0.01%). One might say I’m arguing for a ‘non-negligible probability’ but I dislike the term partly because it seems to come with assumptions about a threshold beyond which probabilities are negligible, which depends on normative and descriptive beliefs.

  2. ^

    This is compared to ~1010 humans, ~1011 farm animals, and ~1018 insects in the present day. Hence, microorganism populations are ~20, 19 and 12 orders of magnitude larger respectively. In addition, in the future, humans may require or desire the construction of new habitats, incidentally creating populations of microorganisms that may experience suffering. Single cells of multicellular organisms may also be considered to be similar to microorganisms, which further increases the total number of entities of the same kind.

  3. ^

    The linked paper estimates median wild bacteria doubling time to be ~7 hours. There are 1030 bacteria in the wild. Assuming equilibrium, a rough calculation suggests that there are ~1.4x1029 bacteria deaths per hour on Earth. This is just a very rough calculation as using the median might not be appropriate and the surveyed bacterial species were limited. If we assume lifetimes of ‘hours to weeks’ noted by the other two articles, calculations could suggest 1027 to 1029 deaths per hour at equilibrium. Of course, bacteria aren’t the only form of microbe, and maybe the true range is somewhat higher.

  4. ^

    However, there may be other factors that determine how much we should care. We might care about the intensity of suffering experienced – although my preliminary understanding is that it’s hard to say whether microorganisms suffer more intensely or less. The problems of fuzzy, nested minds also present difficulties when making comparisons between the suffering associated with different groups of physical systems. See Appendix: Maybe the suffering isn’t intense?

  5. ^

    “Micrometazoans”

  6. ^

    I’m a bit skeptical of this slightly dualistic framing but I think it’s still useful for most purposes. While this is beyond the scope of this post I largely suspect some form of monism is true. I remain confused about the topic.

  7. ^

    Even with other ways of trying to show that others can suffer we can never definitively prove that that’s the case. Our inability to access/be other minds means that we’re in some sense only guessing at what it’s like to be them, if they exist, and never confirm our conclusions to the rigor offered by actually being other minds. This problem applies not only to ‘weirder’ ideas like the possibility of microorganism suffering but also to ‘common sense’ positions like the possibility of suffering of humans other than ourselves. Later, I’ll argue for assigning some credence to microorganism suffering with reference to our credence of more ‘common sense’ positions. Independently proving that others, human or microorganism, can suffer seems impossible in light of the problem of other minds.

  8. ^

    This is not to claim that criteria must be defined prior to examining details of the system. It may be that reflecting on the details of the system is important for forming criteria.

  9. ^

    It might be argued that exploring fundamental physics is an example of something that requires no criteria. Fundamental physics simply is. However, there are criteria, i.e. being fundamental.

  10. ^

    Similar in what ways? Interpreting the system to find features you feel are worth focusing on might also involve a lot of subjectivity. For a list of possible features see this.

  11. ^

    The linked pieces go into this view in a lot more detail.

  12. ^

    If we do have such intuitions, we should be mindful of not dismissing them in bad faith, for example, because we’d prefer microorganism suffering to not exist because it would make our lives inconvenient. On the other hand, if we don’t have such intuitions, we should be mindful of not accepting them simply because of, for example, biases that make us want to accept novel ideas that admittedly might be somewhat intriguing. There’s no reason to think that these biases would help us discover whether microorganism suffering exists, and so we might want to be mindful of them.

  13. ^

    Of course, deciding that uncertainty means we should not be too confident about ruling out alternative views is also a subjective process. However, it’s a decision that seems to make sense. For example, in many other contexts, if we can’t be confident in agreeing to one possibility or some possibilities among a range of possibilities, we would not be able to rule out the alternatives. In addition, if we have nearly as much justification for one position as we have an alternative, being somewhat confident of the first position seems to mean that we should be somewhat confident of the alternative.

  14. ^

    This is related to the possibility of ‘hidden qualia’.

  15. ^

    Brian Tomasik gives the following example in one article: “Ordinarily we might envision higher-order thoughts like this: First-order processing: brain visually identifies a cloud in the sky. Second-order thought: think to yourself, "I see a cloud". But if the cloud is part of your extended mind, then we could reconceptualize the situation like this: First-order processing: cloud moves through the sky. Second-order processing: brain visually identifies a cloud in the sky. Third-order thought: think to yourself, "I see a cloud".”

  16. ^

    Note that microorganisms (as defined) are a diverse category. Prokaryotes consist of bacteria and archaea, while eukaryotes include protists, unicellular algae, and unicellular fungi (e.g. yeasts). It’s estimated that there are around 1030 bacteria, 1029 archaea, 1027 protists, and 1027 fungi. These numbers, combined with facts about these different types of microorganisms, may be useful in deciding for yourself what the expected scale of microorganism suffering might be (especially relative to other sources of possible suffering).

  17. ^

    Macromolecules refer to nucleic acid, proteins, lipids, combinations of those etc.

  18. ^

    A relatively well studied phenomenon possibly due to its relevance to antibiotic development.

  19. ^

    For some of these examples I’m unclear whether they are constitutively ‘always on’ or induced only when potential harm is detected.

  20. ^

    Unfortunately, the tone of the videos might seem a bit too insensitive (if we consider the possibility of microorganism suffering).

  21. ^

    For example, fictional intelligent aliens, many with humanoid features, are often implicitly portrayed to be sentient in that fictional context. We don’t have to be told that they are supposed to be sentient – we just recognize it.

  22. ^

    This isn’t to say that certainty is necessarily justified.

  23. ^

    For example, Buddhists believe in the possibility of reincarnation as non-human animals. Jains note the need to consider plants and microscopic organisms (which they believed in prior to scientific confirmation).

46

5 comments, sorted by Click to highlight new comments since: Today at 11:05 PM
New Comment

We already have a robust scientific understanding of the biochemical causes of bacterial behavior. Why would we posit that some form of cognitive processing of suffering is also involved in controlling their actions? Endorsing a program of study purely on the basis of "important, if true" seems like it would also lead you to endorse studying things like astrology. Since you're mainly taking the fact that bacteria "look like they might be suffering," it seems like you should also be concerned that nonliving structures are potentially suffering. Wouldn't it be painful to be as hot as a star for up to trillions of years?

We already have a robust scientific understanding of the biochemical causes of bacterial behavior. Why would we posit that some form of cognitive processing of suffering is also involved in controlling their actions?

I think even if we have a robust scientific understanding of, e.g., the human brain, we would still think that human suffering exists. I don't think understanding the physical mechanisms behind a particular system means that it can't be associated with a first-person experience.

Endorsing a program of study purely on the basis of "important, if true" seems like it would also lead you to endorse studying things like astrology.

So I'm not looking at the topic only because "important, if true". As I explain in the piece, I think there are signs pointing toward "there's at least a small chance" based on reflecting on our concept of what we take to be evidence of suffering, and aspects of microbes that fit possible extended criteria.

I think there's a lot of scientific evidence against astrology. On the other hand,  how we should attribute mental experiences to physical systems seems to be an open question because of the problem of (not being) other minds.

Since you're mainly taking the fact that bacteria "look like they might be suffering," it seems like you should also be concerned that nonliving structures are potentially suffering. Wouldn't it be painful to be as hot as a star for up to trillions of years?

I am actually concerned that there's some chance various nonliving structures do suffer.  Digital minds are one example.  I don't know much about astronomy/astrophysics but I doubt that stars have functioning that strongly matches, e.g., damage responses exhibited by animals. This isn't to say there's absolutely zero chance we can attribute negative experiences to such a system, but I might think it's much more unlikely than microorganisms that better match damage responses of animals. It might be valuable for someone to look into whether some astronomical objects match some interpretations of what it means to be evidence of suffering, and to what extent. I think there's quite a lot of uncertainty surrounding questions in the philosophy of mind (e.g., some philosophers endorse panpsychism while others might be more skeptical about the sentience of non-human animals).

Bear with me, I can only reply piecemeal.

I think even if we have a robust scientific understanding of, e.g., the human brain, we would still think that human suffering exists. I don't think understanding the physical mechanisms behind a particular system means that it can't be associated with a first-person experience.

On further thought, I think that the most tractable way to pursue your research agenda is to try and discover the mechanistic link between biochemistry and consciousness. It seems to me that this missing link is the main factor that leaves room for possible bacterial suffering. If, as seems likely, scientists conclusively show that either brains or neural architecture are the necessary conditions for conscious experience, then that would heavily downgrade the likelihood that some alternative mechanism exists to provide bacteria with consciousness.

This seems like a tractable angle of attack on the problem, one where success provides benefits no matter where it ends up.

Bear with me, I can only reply piecemeal.

Sure, that's alright :)

If, as seems likely, scientists conclusively show that either brains or neural architecture are the necessary conditions for conscious experience, then that would heavily downgrade the likelihood that some alternative mechanism exists to provide bacteria with consciousness.

One of the things I mention in the post is that whenever we're looking at scientific findings, we're imposing certain standards on what counts as evidence. But it's actually not all that clear how we're supposed to construct these standards in the case of first-person experiences we don't have access to. Brains/neural architectures are categories we invent to put particular instances of "brains" and "neural architectures" in. They're useful in science and medicine but that doesn't mean referring to those categories with those boundaries automatically tells us everything to know about conscious experience/suffering.

What we're really interested in is the category containing systems capable of suffering, and there are a number of different views on what sort of criteria identifies elements in this category: some people follow criteria that suggest only other humans are similar enough to be capable of suffering[1], some people follow criteria that suggest mammals are also similar enough, some people follow criteria that suggest insects are also similar enough. These views have us decrease our threshold for acceptable similarity.[2] One next step might be to extend criteria from the cell-to-cell signaling of nervous systems to the intracellular signaling of microrganisms. If we're confident in/accept some of the other criteria, can we really rule out similar adjacent criteria? This seems difficult given how much uncertainty we have about our reasoning regarding consciousness and suffering[3] . We only need to assign some credence to the views that count some things we know of microbes as evidence of suffering (e.g., chemical-reactive movement, cellular stress responses, how microbes react to predators & associated mechanisms) -- in order to think that microorganism suffering is at least a possibility.

There's a lot of subjective judgment in this. And scientists can't escape it too. The evolution of human intuitions was guided by what's needed for social relationships etc., not fundamental truths about what is conscious/what can suffer. Scientists also impose their standards on what counts as evidence of consciousness/suffering -- but the standards are constructed with those intuitions! I discuss more in the section linked in the footnote. [3] 

discover the mechanistic link between biochemistry and consciousness. It seems to me that this missing link is the main factor that leaves room for possible bacterial suffering.

(I actually I hold a physicalist (monist) view. I don't think consciousness is something additional to physics. Dualism seems unlikely mostly given how science hasn't found additional "consciousness stuff" over and above physical reality. My (very rough) approach to all this is that (some) physical processes (e.g. in the brain) can be said to be conscious experiences, under some interpretation/criteria. Although, I don't mention this stuff in the post to be more theory-neutral (hence my describing things as physical 'evidence of suffering'), which is okay since I think the post doesn't really rely on physicalism being true)

  1. ^

    Similar enough to oneself -- the only system that we're absolutely sure can suffer

  2. ^

    What about more "fundamental" differences? I respond to that in the Counterarguments subsection.

  3. ^

I don't have the time to elaborate, but I find this post compelling