Following The Food Vibes
Can we get even in the neighborhood of the truth without having done all the research?
The legendary architect, engineer, and scientist Buckminster Fuller is credited with describing the concept of the "knowledge doubling curve", a rough estimate that the rate at which human knowledge doubled was accelerating. He theorized that, up until 1900, human knowledge doubled about every one hundred years, but that by the end of the Second World War, the rate at which knowledge doubled increased to roughly once every twenty-five years.
Over a year ago, I was asked for my opinion on the quality of the food available through the United States' industrialized food supply chain. But what do I know? Like many people, I'm a mixed bag of contradictory impulses and intuitions. I have the vague unease that something is going wrong -- health indicators shaped like hockey sticks, scary headlines I have a foggy recollection of, a lot of hue and cry -- but again, what do I actually know? Not much. My attitude could take any direction based on both my whether I overstate how applicable my experience is and by whether I’m generally trusting or distrusting.
I emphasize here that each of these tendencies does not suggest a particular attitudinal direction -- cautious or bullish -- about modern food, but instead cleavages based on partial and over-specialized heuristics.
We have never been able to take it all in. So we survive our information overload by developing instincts that operate as a kind of filter, before our capacity to actually make decisions on the basis of fact. These prejudices help us navigate complicated waters, but has an obvious downside: the world is complicated, and these heuristics, over-specialized and partial as they are, means that we get it wrong all the time. We could presumably fix that by developing a deeper understanding of some topic first. But this supposes that one has the time, energy, and resources to devote to the question. What if you don't?
We need a well-ordered intuition
I think that we ought to spend more time focusing on and developing this general-purpose intuition, rather than relying on our experience alone to develop one for us. I believe this for two reasons: first, because our own specialized experience develops only specialized intuition; and second, because the world is getting complex faster than we can accommodate it.
To the first point, we tend to think that this intuition comes only from the accumulation of many small and specific competencies; that the sum total of our individual strands of expertise and experience can accrete over time a foundation of intuition that can serve us well as a general-purpose tool when we approach unfamiliar territory. If only! The hard answer here is that instincts are quite often contextual. You would never task a successful startup founder (move fast and break things?) to manage the reactor room at a nuclear power plant. A successful teacher in a special-education classroom would be a poor fit as a drill instructor. The instincts and tendencies and reflexes, as effective as they are within their domains, do not translate.
Next, as should be painfully obvious to us, the world is accumulating more and more of these niches or domains where expertise is not easily portable. The marginal value of a general-purpose intuition is therefore increasing! Having a flexible and skill-invariant intuition for complex systems is universally applicable because all systems are complex systems.
You and I encounter these low-information, high-vibe dilemmas every day. So much of the world we interact with is truly beyond our ken. How do we continue to live life both without a quiet intellectual withdrawal and without just being wildly wrong all the time? Food is a useful place to start. We eat it just about every day, it is both literally and conceptually formative to us, and it is an area in which most of us have no relevant specialization (most of our ancestors moved off the farm a few generations ago). What principles might we derive from a careful consideration of food?
1. Take necessary but not suffocating precautions
What might we learn from food? The sheer scale and longevity of functioning ecosystems and their attendant nutritional services inspires awe. It also suggests a principle called the precautionary principle, widely used and known in environmental circles but with powerful applicability generally. In summary, it argues that the burden of proof for whether an action is harmful lies on the actor. If I propose that a new pesticide ought to be used, I must have some kind of certainty that it will not have unintended consequences. This is a deep and profound inversion of the typical approach to this kind of question.
Is the application of this principle always reasonable? Absolutely not. The precautionary principle without moderation would yield a suffocating sclerosis. The more moderate formulation might call for the precautionary principle to take effect in the following conditions:
1. When the systems into which an action is intervening have complex interactions
2. When those complex interactions are not fully understood
3. When the prospective action is irreversible
4. When the prospective action does not leave open optionality but instead precludes other kinds of future action
5. When the systems into which an action is intervening are critical to the continuation of life on earth
How might we judge actions taken in the realm of food production on these grounds? History affords us many examples of cases where these principles, if applied, might have helped. Overfishing, poor topsoil management, deforestation, and the introduction of competitive species were all done with an eye toward improving yields or reducing waste and all had, over the long term, the opposite of the intended effect. But at the very least it ought to temper our expectations about what kinds of interventions lie in the realm of the possible and raise the stakes for when and how we ought to make our own changes.
These principles of restraint, and of times when restraint is reflexively warranted, have application for the judgement of any kind of intervention into a complex system that meet the above criteria. It suggests a way to bridge between specializations, to look for patterns where the innovation or caution appropriate to one niche of our complicated world is applicable to another.
2. Look at the incentives
We should also ask what food is for, in the deepest sense, and therefore why we modify the food system. To put a sharper point to it, we ought to ask who benefits from tinkering with a complex system like the food web. It is certainly true that we modify food to benefit us nutritionally. But given time and resources, what else can our ingenious species pull off? What would be beneficial to us, to them, to their suppliers, or to the exigencies of a global food supply chain? Here are a few suggestions:
Nutritional benefit
Nutrient density
Taste
Drought-tolerance (resource efficiency and resilience against inconsistent precipitation)
Ease of transport (lower proportion of damaged goods, resilient against longer times in the supply chain)
Aesthetics (does it look and feel desirable?)
Pest-tolerance (resiliency against pests, obviation of the need for intensive pest control)
Herbicide-tolerance (resilience against weed killer so that farmers can use more intensive weed control without damaging the produce)
Abundance of yield (more yield per acre)
Size of produce (More yield per unit)
Temperature-tolerance (allowing earlier or later growing seasons)
Growth rate and time to harvest (allowing multiple seasons or squeezing in seasons in latitudes that wouldn't ordinarily have the time for a full cycle)
This is not an exhaustive list, but all of these pressures and optimizations are in effect for the food we eat! For our part, as mostly buyers of food, we make purchases of one food item or another because we like it and we can afford it. Generously, three or four of the optimizations above are ones in support of those immediate needs. We might say that the rest do bear upon affordability. Some, at least, of those savings are passed on to us. But it should be obvious that the motive of profit maximization arises not out of a desire to minimize costs to buyers but to, well, maximize profit. If on the other hand that profit can be maximized by selling overpriced goods to fewer people, that is what will (and does often) happen.
None of the profit-maximizing optimizations found in the food supply chain are prima facie injurious or in opposition to our goals of finding good, cheap food, but they are often incidental to them. In the best case scenario, it is the marketability of the nutritional benefits of a food that make profit and nutrition mutualistic. But at worst, profit-maximizing optimizations gives real, hard incentives for explicitly anti-nutritional activities:
Lying or misleading about the health benefits or risks associated with food, either directly via advertising and messaging or indirectly by exerting financial influence on private research
Regulatory capture to put the thumb on the scale of ostensibly public-interest rules, tipping the schema toward profit maximization. Again, this could be merely incidental to nutritional benefit but removes any safeguard that it is.
Most importantly, making changes to the composition of food for higher yields, better transport, cheaper production, etc., but in ways which knowingly diminish the nutritional benefit of the food
Again, these anti-nutritional activities can and do happen every day. It's hard to say how endemic they are, so I am not making a claim about prevalence but about how we need to broaden the way we think about optimization to understand the possibilities that open up. It’s also fair to say that a many of these malign incentives operate not within the growing stage of food production, but in the processing phase, as taste, shelf life, marketability, transportability, etc., are more easily manipulated industrially.
What do plants want?
I want to pause here to discuss food optimization generally. A naive reading of the points above might give the impression that 'pure', 'raw', 'heirloom', 'unmodified' food is better for us by default. I don't think that's necessarily the case. It's useful to also think about what plants (let's discuss flora alone for the time being) themselves optimize for without human intervention. By selective breeding, planting regimes, etc., humans do put survival pressure on plants to yield outcomes they want, for the myriad reasons discussed above. But what do plants want? They want, as much as it can be called that, to survive to reproductive maturity and to reproduce.
We tinker with this process fundamentally, as we often eat the fruiting parts of plants ourselves – part of the plant’s reproductive system – and therefore take care to modify their composition, size, properties, etc., in ways not necessarily conducive to reproductive success. Of course, sometimes these are mutualistic. One reason fruits are tasty is to incentivize the spread of the seeds contained within the fruits! Without imparting any actual intention to evolutionary selection, we could also say that the optimization incentives and pressures under which plants, in a 'state of nature', operate are also incidental to human nutrition and health. That many plants happen to convey useful nutrients is often 'intentional' in the bare sense that providing good taste and nutrition helps the plant, too. After all, the decaying fruit in fruiting bodies can often provide a seedling its first nutrients, too. But in many other cases it is incidental, and in some cases plants actually develop countermeasures against the wrong sort of consumption by developing irritants, poisons, and toxins to ward us off.
What I am attempting to do here is to put aside this notion that the food we eat has always been selected for our health and nutrition. Both by the concerted effort of many different survival pressures and by the concerted efforts of many humans over a long time period, it sometimes has been. But many of these hard-won changes have been undertaken either indifferent to our nutritional needs or, sometimes, in direct opposition to it.
Can we even untangle complex incentives?
In this reading of complex systems there is a surfeit of statements like 'sometimes yes but sometimes no, it's hard to tell, an intervention is not automatically bad but not automatically good, you have to examine everything individually.' How is someone meant to develop a consistent and general-purpose intuition on the basis of these insights? If anything, this complexity might induce that epistemic retreat I warned against at the outset. Nevertheless, we can at least develop a far more watchful eye to the multifarious incentives shot through complex systems. We tend to look for grand, unifying narratives in complex systems, when the reality is that it is far more competitive and individual. There are a lot of actors, and they're all basically trying to do their own thing. Those things can be positive, neutral, or negative for our purposes, but it's neither any use reading some unifying conspiratorial incentive into the workings of some complex system nor seeing in something like the food supply chain the single and laudable goal of keeping you well-fed.
3. Look for alternatives
One good litmus for whether an intervention is potentially dangerous is to test whether it has made the system more resilient against disruption or more uniformly dependent on a smaller set of factors. This is important for obvious reasons: we're all quite dependent on the global food supply chain, and therefore have a vested interest in its resiliency! I therefore look with concern upon decisions implicit in the design of the global food supply chain to reduce redundancy, eliminate many small suppliers in favor of larger providers with higher margins, eschew labor-intensive but more resilient farming practices for less-resilient but cheaper approaches, and the dependence on low-stock, just-in-time delivery across really long distances. This introduces all kinds of potential disruptions that affect availability and costat every stage.
We have briefly discussed the real human toll of random disruptions in relatively localized food supply chains: famine, starvation, disease, extinction. But when those disruptions occurred, their ripples were rarely felt outside of their milieu. Surely tightening the cords of connectivity in the global food supply chain has also raised the (blessedly small) chance of a catastrophic event in one place resonating harm outward to its region or even the whole globe? Decentralizing a food supply chain is like buying insurance. It impoverishes you while shoring you up against the tail risk of catastrophe. If we are concerned about the potential downsides of other kinds of interventions in the global food supply chain -- perhaps increasing susceptibility to catastrophic diseases, a changing climate, super-resistant pests, or lower water availability -- wouldn't we at least want to insulate the world from the impact of these potential catastrophes by keeping some optionality open? Even if something went catastrophically wrong with one of our novel food experiments, it would be rational to cap the downside and make the tremors of potential catastrophes shake less violently. It is reasonable to believe, for instance, that a war in Ukraine should not cause a worldwide shortage of wheat.
4. Measure how fast things have changed
Another useful principle that it is worth examining is a historical look at 'how it's always been'. We live in profoundly ahistorical times. Can we use the trajectory of history to learn anything about where we ought to go?
There are a huge number of potential mistakes and pitfalls to make on this topic that I see a lot of people make. The primary mistake I see often made is bidirectional. It is either a nostalgia for or chronological snobbery about 'the way things were'. Often, this is pointed at a particular time period. Snobs might point to times before the germ theory of disease was understood and note that things weren't so great! Many people, especially the young and the weak, died in preventable ways. Nutrition was often poor and lacking in basic necessities. Starvation was rampant. Diets were poor and very narrow. It was hardly a wonderful time for foodies.
Nostalgists, on the other hand, might recall the halcyon days of the yeoman farmer, where healthful and nutritious food was grown in abundance on one's own land or swapped with neighbors. In these periods especially the provenance of the food you ate was very well-known, and the incentives of the system were much more aligned with your own nutritional needs and wants.
Both of these views of history are incomplete, both rely on cherry-picked moments and places in time to advance their arguments, and neither really discusses the dynamics at play in what were still complex natural and cultural systems intermingling. They also tend to make unnecessarily normative statements about the proper composition of a diet (or of a lifestyle in general) from their preferred culture and time in history.
The race between external change and adaptation
But if we are to make arguments from history, I think that the far more interesting historical argument is instead about the rate of dietary change, not its composition. We can assume that diets, while profoundly differentiated across time, place, and culture, co-evolved within one ecosystem or culture slowly over time. The metabolism, gut biome, and nutritional needs of humans adapted to the profile of nutrition available to them in their area, and the natural resources and predominant flora and fauna supplying these calories in return adapted to pressures from human culture: implicit selection and breeding and planting, weeding, inculcated taste, soil cultivation, farming and hunting and gathering practices and tactics, co-evolving cultural practices, etc. 'Adapting' here in an evolutionary sense is a relatively violent thing. A lot of people, animals, and plants died before reproducing, or reproduced in smaller generations, because they failed to adapt. And this mediated dance between nature and culture wasn't always mutualistic. Plenty of human interventions were too quick or too wrong to provide adequate conditions for natural systems, and likewise, plenty of sudden natural interventions -- drought, monsoon, disease, weather, fire, famine -- wiped out those humans unlucky enough to rely too heavily on that particular niche of the natural world. Nevertheless, we and nature are still here because enough of our forebears survived this back-and-forth shuffle, which means that adaptation must have been successful in enough contexts.
The important thing here is not necessarily these well-known dynamics, but instead the speed at which they changed. These historical changes were for most of life on earth rate-limited by the number of generations it took for natural and cultural evolution to occur and also by the relatively limited tools available. And if a slight perturbation in this system caused rippling effects, famine or malnutrition or variable resilience against disease or greater adaptability to environmental variations, the fallout was likewise limited and provided a quite brutal input to human co-evolution in the form of mortality or diminished fecundity.
Compare that to what is now the speed of change we modern humans can force. The universal availability of the global supply chain, the ability to do work at the genetic level, and the relative plasticity of matter we have achieved in a short period of time has made those experiments and perturbations literally orders of magnitude faster and more extreme. It has yielded some absolutely stunning results. I'm avoiding normative language here because some of those results -- the fantastically increased yields, advances in plant resilience and drought-tolerance -- are really impressive and have saved an incredible number of lives. On the other hand, cascading species collapse, desertification, and the insane abundance of nitrogen in waterways are unalloyed negatives and came about because of large-scale and irreversible experiments with the global food supply chain. We could accept that man of these large-scale experiments (though absolutely not all) were necessitated by the asymptotic nature of global events -- population increase, or even the quantity of information problem I began with. But a persuasive case can nevertheless be made that the speed and irreversibility of change, not the fact of change itself, is a real problem that is transforming the world into a place inhospitable for our slowly-adapting bodies.
Climate change: is it about warming, or about how fast it’s warming?
Another example of the adaptation problem is climate change. It has been far hotter and wetter than it is now worldwide at other points in the earth's past. But temperatures and their attendant effects have never moved so quickly across the entire globe, ever, except in the case of catastrophic sun-blotting-out events like a meteor strike. Not by a long shot. This means that there is, in world-historical terms, very little time to make the monstrous shifts in the distribution of human populations, huge adaptations in the way ecosystems function, and mind-boggling financial investments needed to deal with all of the fallout of those relatively rapid moves. If temperatures and sea levels rise 2 degrees Celsius and 2 meters over the course of ten thousand or even one thousand years, you would see ecosystems and human populations subtly shifting and adapting over time to accommodate these changes. But those same changes compressed over 100 years run the risk of enormous upheaval and collapse, because there simply isn't time to adapt.
Adaptation is naturally rate-limited by how often organisms reproduce. The rate at which we can tinker with complex systems far outstrips the rate at which the complex biological building blocks within those systems can adapt on their own. In the past 100 years, to pick a fairly arbitrary date in the past, the composition, quality, quantity, and provenance of our food has completely changed. That is a change without precedent.
How fast are we being changed?
And it isn't just food. Novel elements in our soil, water, air, and blood have arisen quickly. (Of course, food or water is often the delivery method for the novel element from soil, water, and air to our blood, so it is fairly influential). Our activity levels and biomechanical patterns of life have changed dramatically and without precedent. Our social milieu has already turned over completely many times over. Perhaps some of us are adapting well to those rapid changes. But there is a compelling case to be made that many are adapting poorly. And perhaps some of those changes are normatively and uncomplicatedly bad. Maybe microplastics in the blood confer no evolutionary advantages for any mammal ever. But some of them are also good! I’m perfectly happy that the polio virus has had trouble adapting to our modern environment. More pessimistically, maybe genetically modified biota will be some of the only species to survive these rapid changes and will therefore form the backbone of some future food supply chain. But the scale and scope of the changes made, as well as the preliminary evidence in various hockey-stick-shaped graphs measuring anything from obesity to neurological disorders to cancers to global temperatures, drives me back to the precautionary principle. Maybe it's all just too fast. Maybe we need more time.
The call for more time, at this point, may just be naive and reactionary. Pointing out how ironic it is that technological conditions which are arguably ending 'life as we know it' are only survivable by further and faster-paced technological advancements achieves only a Pyrrhic, rhetorical victory. In the end, maybe we will have no choice but to 'live in the pod and eat the bugs' because the pod is the only place cool enough, and the bugs are the only protein source that grows anymore.
What do we do?
I'm being unnecessarily pessimistic here. The reality is that rumors of the instability of these nascent system-interrupters have been consistently exaggerated. But tinkering with them increases tail risks significantly. But at the outset of this essay I suggested something far more general and optimistic, which is that we can develop good general-purpose intuitions for complex systems without being subject-matter experts.
The conceit of this, however, is that there is some utility to having the right opinion about some toweringly complex, inscrutable and impenetrable system like food. The nature of these inquiries plumbs far too deep to be immediately useful. What is the matter that a remote and varied set of events and decisions have yielded the food system we have today? My ability to interact with it is so profoundly circumscribed by those far-off and untouchable nodes that developing the mere knowledge of them, without any commensurate agency on my part, can be a psychological torment, a prison made by knowledge combined with an impotent smallness.
But we all eat. And, though circumscribed, our choices are numerous. Within my own remit, how can I follow these principles for interacting with complex systems in my own personal interactions with global food supply chains or, for that matter, global finance, communications, and economics? For my part, I have virtually no direct or indirect control over the decisions to intervene in these complex system. But I do interact with them. I think often of options, backups, complex and multiple interdependencies, and redundancies. How can I order my life and the mutual life of my family such that the overlap between our family's 'food supply chain' and the big one is mediated, contains multiple inputs and outputs, has backups and redundancies, has little independent interdependencies scattered throughout?
Do we really need it all?
More so than that, how are the choices we make couched in our cultural milieu? In other words, can we still have a successful personal food supply chain if we can't derive from it every last ingredient required to make a Big Mac or exotic fruits not native to our region or grown in our season, or can we do without? Have we set our implicit standards so impossibly high because of our exposure to the global food supply chain that we never could (nor should we) recreate those conditions? I live in Florida. Perhaps the near-magical quality of a food supply chain that can bring me instant out-of-season blackberries from around the world is not worth the potential cost, if not in actual harm, than in the tail-risk of low-optionality? No one is going to die because they can’t get blackberries whenever they want, of course, despite my childrens’ claims. So there is no real persistent harm done to us if this supply suddenly shuts off. A more realistic approach might be simply to choose not to participate in experimental interventions in food optimization itself, selecting instead those relatively less modified foods available to me, even if I may be missing on some of those advertised benefits.
The nice thing about the exponential increase in risk that each intervention into a complex system makes is that it can work backwards, too. Even making individual changes at the margins on my own to build that optionality and independence can have a really high, marginal impact. Basically, it's comparatively easier to make yourself half as dependent on the vagaries of the global food supply chain than it is to go the rest of the way and become completely indifferent.
These concerns about the macro-level, being as they are concerns primarily about the way a very large and complex system operates, nevertheless apply to the micro-level. In other words, not only ought we to concern ourselves with the global food system, but also with the interventions into our own biota by the consumption of some food or another. After all, not all macro-level food system interventions are created equal. It stands to reason that, likewise, some of the food produced by the modern industrialized food system is fine or even better, and that some of it is worse or even actively harmful.
As I hinted at before, it's hard to disentangle the health impact of any one particular kind of food from the quantity of it we eat, the proportion of it to other foods we eat, and the opportunity cost of substituting other food we might eat for it instead. Therefore, not only has the method of production and delivery changed, not only has in some cases the very composition or genetic code of food changed, but cultural attitudes and expectations about what to eat and how much of it to eat have also changed. As with the too-big-to-influence decisions about the way the global food system operates, I can only recommend to take appropriate precautions, consider what incentives you might have to eat one food or another, and what incentives brought them to your grocery store (or kitchen), avoid putting all of your literal or figurative nutritional eggs in one basket, and to just slow down how quickly you make changes to that diet.
My hope is that these building blocks of intuition can inaugurate, rather than forgo, our inquiries into any complex topic where we lack the background. Perhaps it is better to think of these principles for developing a general-purpose intuition instead as starter questions. First, we can take appropriate precautions once we know the state of available knowledge in the domain and the potential scale and scope of the impact of making a decision one way or the other. Second, we can examine the multiple incentives embedded in the complex system and reason from them what kind of behavior and counter-behavior we might expect when a decision is made. Third, we can consider whether a particular decision opens up or forecloses independent possibilities. And finally, we can attend to history with a critical eye, neither romanticizing nor dismissing the lessons it teaches.
If you are busy, lazy, and sufficiently exhausted by your own weird specializations like me, I hope that you can find here some principles which, so cloaked in theory as they may now be, will recur in practical form in your next close encounter with a strange and complex system.
Note: After writing this essay and queuing it for publication, I read a wonderful essay by David R. MacIver that in part made an articulate case for the use of analogies (like food?) to think with a topic (like complex systems generally?). You can read it in full below: