DJ Strouse

My Goodreads Rating: 3 of 5 stars

Clearly Braun is not familiar with the recipe for modern pop science texts. Where are the extrapolations from statistically insignificant correlations to bold sermons launching the next consumer craze? Why have they been replaced with tempered, conservative statements accurately reflecting the uncertainty of the scientific process and our current state of knowledge?

Genre-bending accuracy aside, Buzz is a handy user manual for the human body and the two drugs you almost certainly abuse it with – caffeine and alcohol. Braun employs an entertaining, Magic School Bus-style strategy of conveying the science from the point of view of our molecular stars as they journey through your poor unsuspecting body. If you maintain a healthy information diet (or frequently [ab]use either substance), you are unlikely to find many stunning surprises in the discussion of behavioral consequences (Egads! Alcohol disrupts learning and proper sleep and caffeine improves cognitive speed on mundane tasks and is a mild diuretic?!), but the basic science behind their commercial production and effects on the human body offer a few fascinating tidbits:

1) Alcohols are actually a quite large family of molecules. The one you are most well-acquainted with and commonly refer to as “alcohol” is ethanol. However its not the only member of the family capable of getting you drunk. Methanol, just a carbon atom away from ethanol, can also induce intoxication. The reason you do not see methanol on the shelf at your liquor store, however, is that a methanol hangover comes with a slightly less appealing side effect than a mere hangover – permanent blindness. Methanol is broken down into formaldehyde by an enzyme that is found in your retina… and formaldehyde is not something you want your eyeball getting cozy with.

2) That most of the table wine you find weights in at 12% alcohol content is no coincidence; its a necessary condition of the production process. Ethanol is typically produced by the gasping breaths of suffocating yeast cells, and in a 12% ethanol bath, ethanol can no longer diffuse across the yeast cell wall, inducing the drowning cell to shut down.

3) Caffeine, contrary to popular belief, is not exactly brain fuel. It works by blocking the activity of adenosine, an inhibitory neurotransmitter that seems to build up in the body throughout the day. Thus caffeine works by “turnING off the brake” rather than “hitting the accelerator.” This is important because it makes it nigh impossible to overdose on caffeine. On the other hand, this means that if you are a lifeless drone devoid of passion, caffeine cannot rescue you.

One question I leave for researchers of caffeine is: does there exist a biochemical means by which caffeine might make us think or remember that we are/were much smarter under its guidance than we really are/were? Many claim to be granted creative superpowers by caffeine but current research has not been able to support these claims. Perhaps caffeine only increases our beliefs about our cognitive abilities and not our abilities per se.

View all my Goodreads reviews

My rating: 4 of 5 stars

Book Review
Never have I read a book that communicates the process and logic of scientific discovery so well. Like erotic literature for the scientist, Wisdom of the Hive not only conveys what entomologists know about bee colonies but the graphic details of they found out. Seeley prefaces every discussion of experimental data with the precise thought process that led him or colleagues to perform the experiments as well as a clear overview of all methods used. He follows each discussion with scandalously honest assessments of what can and cannot be concluded from the results. He even has the grit to discuss competing hypotheses (i.e. views he does not hold), past and present misconceptions in his field (i.e. times he and others were wrong), unresolved problems (i.e. stuff he hasn’t figured out), and suggestions for future experiments to resolve these problems (i.e. ideas he has not yet had time to pursue and that could be taken up by others). Perhaps most importantly, Seeley has the discipline to not blow his scientific load early and lead discussions with experimental results and conclusions. Instead, he carefully walked me through the historical results and thought process that lead to a particular question, considered possible routes to resolve this question, and only then revealed that: “Oh by the way, I’ve performed this experiment and here are the results and how I interpret them.” In other words, Seeley never answered a question I didn’t have; he takes careful steps to ensure that I was practically begging for the answer when he presented it. The only danger in going into such detail is that Seeley has to spend the first four chapters and eighty pages introducing the reader to bee physiology, experimental methods in entomology, and the broad topics covered in the ensuing discussions of experiments. I was sipping from these initial pages like a forager bee from a 2.5 mol/L sugar solution feeder after a week-long thunderstorm, but those not sharing my enthusiasm should take note – the book really shines from Chapter 5 onward.

Despite the focus on experiments, Seeley also paints a coherent theoretical picture over all by emphasizing abstract principles of information flow within a hive. Thus, despite the dozens of experiments mentioned and the dazzling complexity of the beehive, I feel confident that I could take up a summer internship in a beehive and never break decorum. He also includes a summary at the end of each chapter to highlight the most important experimental results and open questions. Every field needs a Seeley – someone to provide a comprehensive and even-handed review of methods, past experimental efforts, current agreed upon and disputed hypotheses and models, open questions, and suggestions for future research directions and experiments.

This masterful work can be read as a comprehensive review of information flow in bee colonies, a how-to guide for designing and carrying out experiments, or a near-perfect example of scientific writing for a general audience.

What I Learned
Despite several endeavors into the complexity and chaos literature, I’ve never encountered a better treatise on how global organization emerges from local interactions. Bee colonies elegantly optimize the allocation of labor and collection of resources to satisfy current and projected needs even though colony resource levels and needs are neither known to any single bee nor readily available in a centralized signal. Instead, individual forager bees integrate information about their colony’s needs with the profitability of resources they have discovered, and if the resource is judged important by that bee, the bees performs a “waggle dance” to recruit other bees to join him in foraging from his discovered source. The details of the waggle dance indicate the location of the resource while the duration of the dance is a measure of how important the bee thinks his discovery is to the colony. Since other bees sample dances unbiased, longer dances result in more bees recruited. No Department of Labor or managerial staff – just individual, information-processing, dancing bees. Foragers can also regulate their personal foraging vigor to increase or decrease resource collection as well. (Why not go all out all the time? Its not energetically efficient to do so, and energy seems to be a constraining resource in bee colonies. There is no bee McDonalds or manufacturer of bee Oreos.) The emerging picture is this: if you want to design a complex and powerful organization in which individual members possess as little information on the actions and goals of the organization as possible, the US government a bee colony would be an excellent model.

How do foragers determine their colonies’ needs? Again through local mechanisms – the search time for a processor bee to unload their delivery (in the cases of nectar, pollen, and water) and personal level of protein (in the case of pollen). Short unload time for nectar? Clearly not enough nectar is being collected. Dance a waggle dance to recruit more foragers. Long unload time for nectar? Clearly the processors need to ante up. Dance a tremble dance to recruit more processors. Sustained success of nectar foraging? Clearly the black locust trees are in full bloom. Perform a shaking signal to recruit more foragers. Surplus of protein in the diet for a pollen forager? Clearly the colony has plenty of pollen. Cease pollen foraging and go check out the waggle dance floor to see what the colony really needs.

These mechanisms also introduce the distinction between cues and signals. A signal is produced explicitly to communicate information, while a cue is a byproduct that may act to communicate information. Search times and protein in the diet are both cues while waggle and tremble dances are signals. Why would bees use cues? One reason is that they are easier to evolve. A cue requires only the evolution of a recognition mechanism for an exiting observable instead of the co-evolution of signal production and recognition. There are also cases in which signals would be difficult and expensive, such as employing a bee to survey the colony’s entire resource stores and broadcasting his findings. Why then do bees also use signals? For some information, there does not exist a cue. A returning forager loaded with nectar may be adorned with the scent of flowers which provide some information about his collection source, but the direction of these flowers is not encoded in him in any way. Thus, to recruit more foragers to a profitable source, an explicit signal (the waggle dance) is required.

Colonies also exhibit the influence of resource requirement variability on collection mechanisms and the differences when that variability is supply-driven vs. demand-driven. Because nectar and pollen availability are highly variable, bee colonies do not send all foragers to optimal collection sites but instead distribute them among non-optimal sites as well. This provides for the continual monitoring of resource sites and robustness against rapid shifts in supply. Also, since the variabilities in need for nectar and pollen are supply-driven, bees maintain stores of these resources in their hives. The variability in need for water, on the other hand, is demand-driven, and bees do not store water but merely collect it when needed.

Colonies are also capable of integrating external and internal signals to make decisions. High nectar availability (external) and nearly full combs (internal)? Clearly the colony is running out of space for honey storage. Build more combs. (By the way, how do processors detect comb fullness? Though results were not conclusive at the time of this book’s writing, probably long search times for empty comb.)

Colonies also employ combinations of mechanisms acting on various timescales to regulate their function. Pollen foraging is regulated both by the collection rate per forager (short) and the total number of pollen foragers (long). Why two mechanisms? The former is faster to adjust but provides less dynamic range, while the latter is slower to adjust but provides more dynamic range. The result is a rapid and robust combination of mechanisms allowing colonies to match pollen collection rate to pollen demand and supply.

The above also highlights evolution’s ingenious reuse of biological design principles: the use of search times in nectar, water, and pollen collection to indicate balance between colony demand and supply and the use of dances for communication (waggle and tremble) of resource needs and locations.

In closing, the above language I use is not accidental but is meant to suggest analogy with another system whose investigators might benefit from considering the design principles of bee colonies and the experimental techniques and theoretical concepts of its researchers. That system is the human brain. For those who listen carefully, discussions of global organization implemented by local interactions, the dual use of cues and signals, the essential role of variability, the integration of external and internal signals, the interaction of mechanisms acting on various timescales, the distributed storage of information, the use of excitatory and inhibitory feedback, and the elegant reuse of mechanisms should sound eerily familiar.

View all my reviews on Goodreads