Redundant parts

December 5, 2010 • 9:51 am

My Ph.D. advisor Dick Lewontin recently had a triple-bypass operation (it went well and he’s fine), and during my check-in call he described the procedure.  Like all bypass operations, veins or arteries from other parts of the body (usually the arm, leg, or chest) are harvested to circumvent the blocked coronary arteries.  Dick noted that the body is full of superfluous blood vessels.

“Superfluous?” I said.  “Maybe we can live without them, but couldn’t they have evolved as ‘backups’—in case something went wrong with a partner vessel?”

“Oh, Jerry,” said Dick.  “Don’t be such a selectionist.  Maybe they’re truly redundant: they didn’t have to evolve as backups.”  (Dick, remember, was the second author of the famous “Spandrels of San Marco” paper, arguing that much of biological form was not directly adaptive, but present for other reasons.)

Thinking about it, I think there’s a lot of truth in what he said.  First of all, it’s unlikely that we have superfluous veins because our ancestors had vein problems and needed backup vessels.  What killed them was not thrombosis, but disease, predation, infection, and hard living.

(Caveat: because modern humans can do “fine” without a major artery in the leg doesn’t mean that not having it in our ancestors might have conferred some marginal disadvantage.)

And it’s not just blood vessels that seem superfluous.  We have two kidneys, and can do very well without one.  Do we have two because one could serve as a backup for the other? And if that’s why we have two kidneys and two lungs and two eyes and two testes/ovaries, why only one liver and one heart?

We are bilaterally symmetrical, bipedal organisms descended from bilaterally symmetrical fishes. In some cases having two of something is useful.  Our two eyes give us binocular vision, but we have two not because of that facility, but because our fishy ancestors had two eyes that enabled them to see, nonbinocularly, on both sides of their bodies.  It may be the case that when some new organs evolve, like a lung or kidney or ovary, the genes that make it act bilaterally, giving us two organs instead of one. That, and not an evolved redundancy, may explain why we have “superfluous” bits.  The fact that one can act as a backup is simply a spandrel–a fortuitous but nonselected byproduct of a developmental imperative.

When we ask, then, why some bits are present in pairs and others not, the answer, then, may not always be because it’s better to have two than one.


Coda: The bypass operation has been one of medicine’s great advances, giving many people many years of added, productive life. I remember the first one, which was only in 1960; and for years afterwards it was a dicey and dangerous procedure.  Now it’s more or less routine with a high expectation of success: as Dick said, his hospital was running a “bypass factory.”

60 thoughts on “Redundant parts

  1. “…why only one liver and one heart?”

    Indeed, but I suppose you could argue that the heart is bilateral, with two pairs of chambers.

    1. In mammals, yes, but it’s much more simple in amphibians or fish. Also, it’s not located on the line of symmetry- it’s offset.

      1. The heart quite clearly is bilateral in origin, fusing near the midline as development proceeds. A classic experiment on chick embryos (easily replicated by undergraduates) produces paired tubular hearts, each beating independently.

  2. I think the symmetry of our bodies is another great line of evidence against creationism. I wonder what Gods’ sales-zombies would have to say about it. I hope I remember to ask them next time they come aknockin’, “How come there is symmetry to our bodies if we are created?” Just one more monkey wrench to throw at them and hopefully get their brain juices flowing again.

    1. You’re forgetting that we are created in god’s image, and he clearly manifests as human. It’s in the bible. 😉

      1. Indeed, the Bible is positively brimming with get-out clauses, you just have be dishonest enough to look for them and twist them into what you want to hear.

  3. But symmetry of gene expression doesn’t explain the (variable) excess of upper and lower limb superficial veins.

    1. It seems highly unlikely that the detailed branching structure of the circulatory system is specified explicitly in the genome. My guess is rather that genes set up a process by which blood vessels grow and branch semi-randomly in response to contextual cues, and individual branches either end up doing something semi-useful or else wither away during development. Under that kind of process, the existence of redundant or superfluous veins is not terribly surprising.

  4. Your adviser has a couple of months of feeling like a infant and freezing to death in warm rooms to look forward to. I wonder what it is about a heart/lung machine that makes one’s internal thermostat go wacky.

  5. They are now doing some bypass surgeries laparoscopically. I think that’s great! Hardly more invasive than installing a stent.

    1. bigjohn756, I had a stent placed in 2005 and the placement of it caused my heart to stop during the procedure (for which I was awake). They had to shock me to restart my heart. I think this is a very invasive procedure.

      1. I don’t think anybody would argue that laparoscopy is a non-invasive procedure, or that installing a stent isn’t invasive.

        But in comparison to having your pectoral muscles sliced clean through, your sternum cracked in two, and your ribs spread open to the sky like an Aztec sacrificial victim?

        Let’s not lose perspective. There’s almost nothing more invasive than open heart surgery. If we could cure heart disease by taking people out back and beating them with rubber hoses, it’d still be a vast improvement.



  6. I’ve come to the (not very surprising) conclusion that English is a very poor language to discuss evolutionary topics in. English, by its structure and grammar, tends to presuppose agency, causation and teleology.

    The correct question perhaps is not ‘why do we have spare veins?’ (implied cause and effect) but ‘what set of blind processes and states over time have the consequence that we have as many veins (arteries etc.) as we currently do?’

    1. Well DJ, your comment resonates strongly with me. I’m inclined, however, to lay the blame with English speaking scientists rather than with the language. There are, after all, any number of intransitive verbs to choose from when describing evolutionary change without relying on perpetuating past mistakes with transitive verbs—”selection” being the most inappropriate of all.

      Darwin initiated the error but, to his credit, knew he was making a mistake but underestimated the consequences. I thought that Lewontin’s comment to Jerry Coyne, “Oh Jerry … don’t be such a selectionist” was right on the mark. Our language certainly resists efforts to wring causality and teleology out of evolutionary topics. I think, however, that the bigger problem is the lack of desire. Warm regards, Bill W.

  7. I think a redundant kidney or lung is not something that would be particularly selected for. Until recently (because of technology,) if you had a kidney failure or if one lung failed you’d be so injured or sick that it would probably not matter if you had a spare; you’d probably be on your way to pushing up daisies.

  8. They are not superfluous blood vessels. I had a triple-bypass operation 8 years ago and they took one from my forearm and one from my lower leg. I have numb places quite often in my ankle and a couple of toes and occasionally on my index finger. After querying my cardiologist with concern, he said this is quite common. I had tests run anyway for checking for blood flow (CAT scan and ultrasound) and there is no other reason for the symptoms.

  9. I wouldn’t agree that we can do well with only 1 kidney. As we age, the glomerular filtration rate typically declines and some (many?) people would suffer if they only had 1 kidney. If you’re going to cut out bits of some organ, make it part of the liver – unless of course the donor is already suffering a bad case of hepatic cirrhosis.

    1. Indeed. A friend of mine in his late 70s had one of his kidneys removed. For cancer. He’s now suffered three mini strokes over the past year. These were caused, so I’m told, by spikes in his blood pressure caused by the failure of his remaining kidney to properly filter his blood.

    2. I suspect the organs are dimensioned after total capacity anyway. (Modulo the possible extra surplus of tissue, as noted on the vascular system.)

      The question is: why do we have one liver? I get the one stomach, it’s a tubular modification; as opposed to lungs, which are dead end (and so non-efficient, sigh) organs. Maybe the liver is a similar modification, but on a vascular tube?

      1. A few thoughts:
        Blood vessels may be redundant for the function of delivering blood to and draining blood from various tissues at (relative) rest, but this is not their only function. Bloodflow patterns vary constantly as arterioles and precapillary sphincters dilate and constrict in response to local chemical conditions. On a larger scale, flow patterns to organs is very different when exercising than when resting. And as John Hawks notes above, there is another level of control that regulates peripheral and deep circulation routes for heat loss or retention. Redundancy of any single vessel is a very premature conclusion even if it can be removed without serious-seeming consequences.

        There are some clear examples of backup redundancy; for example, the arterial supply to the brain is connected such that either carotid artery can supply the entire brain if need be.

        As for paired organs, just having 2 does not make them redundant. If one human kidney is enough, then selection could have simply halved the capacity of each, making both necessary. It’s the apparently excessive functional capacity of each one that suggests there is a safety margin evolved in.

        The liver (and pancreas) are single organs presumably because they develop, and always have, from the (unpaired) digestive tract. (And but the same is true of lungs…)

        1. Wait! You mean the measure of whether something is redundant is NOT whether you can remove it from a 75 year old US male (who makes his living by thinking) and it doesn’t disturb his ability to rest in bed?

          I suppose next you are going to tell me that there is an advantage to having *TWO* legs and *TWO* arms, even though I can point you to many people who seem (to me) to get by just fine with one (and some even none!).

  10. I have vacillated on this this topic. For example in 1988 (BioEssays 8:204) I suggested that, given complex combinatorial regulatory mechanisms, selection on one aspect of gene expression could drive gratuitous (or even mildly deleterious) expression in other times and places. Ten years later (PNAS 95:253)we showed that many yeast knockout mutations that had no readily discerned phenotype in fact had small fitness deficits when the the measurements were sufficiently sensitive. I suspect both kinds of explanations are true in various cases, but that it will be difficult in practice to determine which is more important in a any given case.

  11. Glad to hear that Lewontin is doing well. He’s one of the most versatile and original thinkers around. He advised one of my graduate advisors as well. He has trained and influenced so many so positively for so long.

  12. If you’ve ever tried to design a robust computer system — and, I’m sure, much of anything else such as an airplane or a nuclear power plant or a heart / lung machine — then you should know that simply throwing redundancy at the problem is almost guaranteed to reduce robustness, rather than enhance it.

    The classic novice mistrake with computers is switching to a RAID array. The problem, of course, is that most RAID configurations available to consumers require all disks to be healthy in order to operate. If you have a three-disk array, you’re now actually three times more likely to experience a disastrous failure than if you had a single disk.

    Similarly, people think that more engines in an airplane is a good thing. Many people who’d be terrified of flying over the Rockies in a Cessna Caravan (because it only has a single engine) wouldn’t think twice about making the trip in a Beechcraft Baron because it’s a twin-engine plane. Yet the Baron has a single-engine service ceiling of only 7,400 feet. You’re twice as safe making the trip in the Caravan: in either one, if an engine goes out, you’re going down fast…but, in the Baron, your chances of losing an engine are twice as high as in the Caravan.

    (By the way, both are superlative aircraft. In either one, if it’s been maintained according to regulations and if the pilot is sober and current with all certifications, you’re far safer in the air than you ever will be driving on the ground.)

    That’s not to claim that redundancy is bad, of course. With computers, a robust system will, indeed, have multiple disks; they’ll just be configured so that you can lose a number of them and the system keeps chugging along (while telling everybody that there’s a problem that needs fixing, of course). In aircraft…well, the big jets have so much power that, even on a single engine, they still easily outperform general aviation aircraft. A 737 with one of its engines turned off will fly faster, higher, and farther than a Baron in perfect condition will with both engines.

    The general tactic in enhancing robustness is to reduce single points of failure. A naïve RAID array fails, since you’re adding more failure points. An intelligent, well-engineered (and expensive!) system will have multiple independent disks, power supplies, controllers, and more. It may even be designed such that a nuclear war could take out an entire continent yet surviving systems on other continents continue uninterrupted.

    I don’t know, but I suspect that at least some of our own redundancies are of the bad kind. Having two breasts doubles the chances of breast cancer, and having two testicles doubles the chances of testicular cancer. Being able to nurse two infants simultaneously probably outweighs the added risk, from an evolutionary perspective. But couldn’t us guys have had a better arrangement? Not dangling in the breeze would have been nice, for starters….



    1. True; last week I heard a computer being ooh-ed and aah-ed on, and the biggest thing was “it has a raid”.

      Great, now you don’t need time to do backup, you need time to get something to backup in the first place. [Yes, I snickered.]

    2. Statistically, you are correct, Ben, but in practical terms, I like RAID. Maybe it’s because I’m the one who has to fix ’em, but I wouldn’t consider a mission critical server without some form of hard drive redundancy. It’s still a lot easier to slap in a new HD and let it rebuild than it is to completely reload with angry users breathing down your neck. Not a fan of flying over the ocean in a single engine aircraft either. I also bring 2 guitars to every gig, so maybe I’m just anal.

      1. Well, obviously you’re not one of those naïve RAID users I was referring to. If you can pull a drive (preferably more than one) without sacrificing uptime (and, ideally performance), then you’ve achieved robustness through redundancy. Sure, you’ll be replacing drives more often, but your systems will go down less often. You’re Doing It Right™.

        Compare that to the poor gamer who thinks that his l33t RAID-0 array means he doesn’t have to make backups any more, or the almost-as-clueless one who thinks the same of his proprietary software-based RAID-10 array or many poorly-considered in-the-wild RAID-5 arrays with funky controllers….

        Us classical trumpeters usually wind up bringing lots of horns to gigs, but not for backup purposes. There’s the C, which is the standard horn for orchestral work, and the Bb which is standard for pops, jazz, and brass quintets. There’s the piccolo for most everything from the Baroque era plus a fair share of modern stuff (including quintet). Jazz gigs may well call for flügelhorn (which is the one horn missing from my collection), and all sorts of gigs might call for cornet. Some classical works call for Eb or D trumpet. (There are other exotic trumpets out there, but you almost never even see them, let alone have a use for them, let alone actually need them.)

        My standard “grab and go without thinking about it” case has in it the Bb, C, and picc (and a bunch of mutes, a stand light, valve oil, pencils, etc., etc., etc.). The Bb and C are somewhat interchangeable, but you never really think of the one as a backup for the other (until disaster strikes). In many cases, if you have a choice between borrowing another Bb (or C) instead of using your C (or Bb), you’d borrow a horn in the same pitch.

        In reality…I don’t think I’ve ever had an instrument fail me, and I can’t remember the last time I had a gig where somebody had an instrument fail. The closest I can think of is a friend who plays in a brass band…this past summer, the stage hand was bringing the (big-name international) soloist’s gold-plated instruments out on stage, tripped, fell, and severely damaged a couple of them. The soloist borrowed my friend’s cornet for the concert. Moral of the story: never let anybody else handle your instruments, period. (Of course, I make exceptions for other trumpeters I trust if they want to try them.) I don’t even leave them unattended on stage amongst other musicians. The Bb is a pre-war Besson and the C is an early-40s Conn previously owned by two legendary principal trumpeters of the New York Philharmonic (Harry Glantz and Bill Vacchiano), so they really are irreplaceable….



          1. My dad used to be a freelance trumpeter in the New York area. He played the offstage solo on Stokowski’s Symphony of the Air recording of Pines of Rome, he was principal trumpet for both the Bolshoi and Royal ballet tours of the states, and did all sorts of other typical New York freelancer gigs.

            Dad bought the horn from a guy in Florida named Jack Pinto, who bought it from Vacchiano, who had bought it from Glantz.

            It’s not a perfect instrument; there’re a couple intonation problems that’re annoying. But it’s a perfect fit for me (probably in no small part because I’ve been playing it for a few decades, now).

        1. Looks like you need a baroque trumpet, or even a no-holes natural trumpet. And a full complement of crooks. Weihnachtsotatorium performances are approaching!;)

          1. I’ve thought about it. I might even go to one of those seminars where you learn how to make your own.

            On the one hand, Bach is hard enough on modern instruments. I did the Christmas Oratorio a couple years ago with the Scottsdale Baroque Orchestra. (I think that might have been the same season I did the Michael Haydn concerto with them.) It’s a lot of fun, but a hell of a lot of work!

            On the other hand…the best recording I’ve ever heard of the Christmas Oratorio, and of much of the rest of Bach’s works, is Crispian Steele-Perkins’s. It just ain’t fair!

            Of course, once you go down that road, you need both Baroque and Classical natural trumpets, and a cornetto or two, and a keyed bugle for the Joseph Haydn, and…

            …and, before long, you might as well audition for the Chestnut Brass Company and be done with it!

            On an almost entirely unrelated note…the best natural trumpet I’ve ever fooled around with is the Besson with all the slides extended to their maximum length (and, obviously, all valves down). It just happens to put it in D.



            1. Steele-Perkins is a formidable talent.

              I hadn’t heard of the Chestnut Brass. Serpents, sackbuts, ophicleides? Well, let me know when they get their hands on a gross-bass pommer. Then I’ll be impressed.

              But seriously. Good grief.
              Do they play them all?

            2. Amen to Steele-Perkins’s awsomeitude!

              Chestnut did a concert several weeks ago at the brand-new Musical Instrument Museum in north Phoenix. Chestnut’s collection of instruments is at least as extensive as the Museum’s.

              They gave a good concert. It was mostly standard brass quintet fare, except they used period-appropriate instruments. That included cornets, sackbuts, and serpents for early music all the way up through over-the-shoulder Civil-War marching brass (and, of course, modern instruments for the contemporary stuff). There were keyed bugles and serpents, things with weird valve mechanisms that I only remember from history classes, natural horn (of course), slide trumpets (not soprano trombones!), and more, I’m sure.

              They’re no Canadian Brass, but they deserved a far bigger audience than they got.

              They’re strictly brass musicians, so no reed instruments. Hand a gross-bass pommer to a tubist and he’s likely to use it to prop up the hood of his car. But they did have a representative of every member of the serpent family….



            3. Oops. My mistake. Somehow I’d gotten it in my head that pommers were old brass instruments. Probably because the (rare) organ stop “Pommer” is not a reed. And even period instrument ensemble aficionados don’t come across the term very often. Anyway, enough thread derailment. 🙂

        2. Unlike the human body, these things are designed.

          I don’t think it likely that a trumpet will break at a gig, so I don’t think you’d need two Bbs. You bring a nice variety. I have also never lost a guitar or an amp at a gig, but I have broken strings on occasion, necessitating an immediate swap (and blown fuses in the old days before I understood output impedance). I used to bring virtually identical guitars just in case, but now I tend to bring a variety for different timbres. I’m also not has hard on them as I used to be.

    3. It’s been my experience of the computer industry that (paradoxically) any fully redundant system still has a single point of failure.

      The mains power to one of our computer halls failed and our uninterruptible power supply (battery and generator backed) failed to switch in because a 3 amp fuse blew.

      1. I’m pretty sure there’s a datacenter in a nondescript warehouse somewhere here in Phoenix which gets mains supplied from both of the regional power companies, SRP and APS. And, of course, some datacenters not only have multiple battery backups but multiple generator backups and filling station-sized fuel reserves for said generators. Of course, network connectivity would be redundant through each of the regional top-tier carriers, as well.

        Obviously, a disaster — plane crash, bombing, nuclear war, whatever — could still take out the entire complex, which is why banks and other similarly-sized institutions have multiple datacenters scattered all over the place….

        Me? I have a fire-proof hard drive attached to an Airport base station in a closet on the other side of the house which I use for TimeMachine backups. Anything that would take that out would make the data loss look trivial by comparison. Uptime I couldn’t give a flying leap about, except for when I’m on the clock.



  13. Jerry: “Our two eyes give us binocular vision, but we have two not because of that facility, but because our fishy ancestors had two eyes that enabled them to see, nonbinocularly, on both sides of their bodies.”

    Plenty of mammals still have their eyes located on the sides of their heads. I don’t thank anyone is saying the need for binocular vision was what produced a set of eyes. If there was selection maybe it was in the migration to the front of the face, and mainly in predators.

    1. However, all of those animals evolved from ancestors that already had binocular vision.

      We’re all descended from the first bilateria. Eyes on both sides of the midline comes with the package.

  14. Fifteen years ago I had an accident while driving a farm tractor. My left calf muscles were torn down from the back of the knee and I lost a vein that comes from the lower, outside of the left foot. This vein crosses over the lower leg to meet the Great Saphenous Vein just inside the left knee. From the same area of the lower outside of the left foot the Posterior Tibial Vein leads straight up the left side of the foot and was not injured. The calf muscles were reconstructed to allow me to walk without a limp but I developed a postoperative infection in the Great Saphenous Vein. Bacteria migrated up the vein creating boil like pustules along its path. The vein had to be excavated from the inside of my thigh for about eight inches. I asked this wonderful Doctor (Kevin Yokuboff) of the University of Cincinnati Hospital what losing the vein would mean. Tongue in cheek he said, ‘Well if you have another accident and I need a piece of vein I can’t use that one.’ But seriously he said, ‘You are going to have to walk every day to keep circulation going up from the left side of your foot.’ For some time I experienced numbness in that area but at times would get strange feelings in my leg; sort of an oogy itch. Maybe the Posterior Tibial Vein picked up some capillaries from the area and now 15 years later, retired at 83 years old, briskly walking three miles every morning, I have no numbness in that area of the foot. After reading your Why Evolution is True and The Greatest Show on Earth of Dawkins about how the ends of developing nerves, arteries know how to construct a new being, can those genes still know how to construct new paths? They daily continue to construct new skin etc.
    Gayle Stone

  15. The vessels almost certainly aren’t redundant. If you take a reasonably-sized vessel out of the body (as in coronary artery bypass surgery), the body is forced to grow, by angiogenesis, a whole set of new collateral vessels to transport the blood instead. (The result isn’t pretty – you can see a nice example of the squiggly little collaterals in <A HREF=""this angiogram, or in <A HREF=""this schematic).

    I think this strongly argues against redundancy.

    1. Sorry, that should read:

      The vessels almost certainly aren’t redundant. If you take a reasonably-sized vessel out of the body (as in coronary artery bypass surgery), the body is forced to grow, by angiogenesis, a whole set of new collateral vessels to transport the blood instead. (The result isn’t pretty – you can see a nice example of the squiggly little collaterals in this angiogram, or in this schematic).

      I think this strongly argues against redundancy.

      1. Fascinating! Looks like they’ll be growing along, run into some sort of “road block,” and have to turn a corner!

  16. ““Oh, Jerry,” said Dick. “Don’t be such a selectionist.”

    that’s a good one line summary of Lewontin.

  17. (Dick, remember, was the second author of the famous “Spandrels of San Marco” paper

    I’d add that both he and Gould did a remarkably poor job of responding to the legitimate criticisms of that paper.

    but then, I’ve always wondered who their intended audience really was.

  18. Back when I was in the daily journalism grinder, I got to watch a heart bypass operation. I was standing gowned up directly over the head of the patient, looking straight down into the open chest.

    The anesthesiologist advised me that if I was going to be sick, I should do so somewhere other than the chest cavity.

    No worries there: it was fascinating.

    However, the overall impression I had was of it being a somewhat complex and slithery plumbing job.

    I was also impressed with how much fat was around the patient’s heart.

  19. In Bully for Brontosaurus, Gould relays an apt remark from Francis Crick: “Why do you evolutionists always try to identify the value of something before you know how it is made?”

  20. Punctuated equilibrium! Periods of rapid change with a lot of use of preadaptations + quick and dirty application to low-hanging fruit, with fast diffusion of novel traits; followed by long periods of stability with deep valleys between local optima in mutual tension, a rigid web of ecological niches. Innovation becomes fine-tuning, slows to a crawl. Until something happens, some external driving force kicks in.

    There’s a lot of random junk left over from any period of reckless prosperity. The actively unhealthy stuff will die out. Stuff that is neither harmful nor useful will have some kind of half-life, persisting until eliminated as part of some urban redevelopment project.

    After a “long” period of stability, remaining forms are not likely to be new innovations and are not likely to have persisted by chance; they are then likely to be locally adaptive in the sense that fitness would be locally reduced by any local change.

    I presume these circulatory system features are generalized-mammal stuff? So my guess is they are likely to be adaptive. The circulatory system is the Ho Chi Minh Trail, not Interstate 5. Amsterdam hydraulics, not Tombigbee.

    Not an expert on any of this stuff. Feel free to correct me.

  21. You really make it appear really easy with your presentation but I find this topic to be actually one thing which I think I would never understand. It kind of feels too complex and very huge for me. I’m having a look ahead in your next put up, I’ll try to get the dangle of it!

Leave a Reply