Home » Intelligent Design » MIT’s Department of Biological Engineering

MIT’s Department of Biological Engineering

Here’s how MIT describes its department of biological engineering. Does the research here fall more readily under ID or Darwinism?

Biological Engineering [BE] was founded in 1998 as a new MIT departmental academic unit, with the mission of defining and establishing a new discipline fusing molecular life sciences with engineering. The goal of our biological engineering discipline, Course 20, is to advance fundamental understanding of how biological systems operate and to develop effective biology-based technologies for applications across a wide spectrum of societal needs including breakthroughs in diagnosis, treatment, and prevention of disease, in design of novel materials, devices, and processes, and in enhancing environmental health. The innovative educational programs created by BE reflect this emphasis on integrating molecular and cellular biosciences with a quantitative, systems-oriented engineering analysis and synthesis approach, offering opportunities at the undergraduate level for the SB degree in Biological Engineering, and at the graduate level for the Ph.D. in Biological Engineering (with emphasis in either Applied Biosciences or Bioengineering).

SOURCE: MIT-Biological-Engineering

  • Delicious
  • Facebook
  • Reddit
  • StumbleUpon
  • Twitter
  • RSS Feed

33 Responses to MIT’s Department of Biological Engineering

  1. I’ve seen posts like these before. I’m wondering if this means some see a new direction for ID – not only learning from natural designs (And assuming that nature, both as a whole and in part, is essentially a design), but also advancing their understand, admiration, and even further use?

    That’s what I’m hopeful of, at least.

  2. After browsing the course names, I’m not seeing a course on evolution. Look at the Biology department as well.

    When trying to make engineering solutions by looking to nature, we should notice that “make random changes and see what works” is never offered. Even though that scenario requires an intelligent agent to do the work of natural selection. But we don’t even get that.

    Random mutation as an engineering solution is a non-starter. There is a reason for that.

  3. It most definitely falls under ID. An central theory in ID is design identification (i.e. Explanatory Filter).

    Furthermore, there are numerous designs which can be mathematically demonstrated to be poorly identified through the presumption of natural selection. The examples that comes to mind are those designs that are of the back-up system variety (in automobiles, spare tires are such designs).

    You can knock back-up designs out, and the organism is just as reproductively fit if not more so (because of metabolic reasons).

    Darwinism will fail to detect and identify such designs, but the explantory filter (which all engineers use implicity) will be far more effective at identifying such designs.

    Like what happened in the medical schools in the USA, I predict will happen here at MIT, the students will be a bit resentful if Darwinism is shoved down their throat since Darwinist dogma will be quite useless to systems analysis.

  4. Interesting point, Sal. To promote ID, should we encourage Darwinists to make Darwinian evolution a required course for engineers? And having been force-fed Darwinism for a semester, are they likely to resent this intrusion as an exercise in futility that does not in any way further their ability to conceive and implement engineering designs? As Phil Johnson always said, it is in our interest to teach not less but more evolution (though his intention in saying this was that Darwinian theory should be taught honestly — warts and all).

  5. Bill,
    Would you have the interest/inclination to and/or Is there a chance you could offer up time to do a lecture there on ID and how it’s relevence?
    The exposure to ID might be a worthwhile seed to plant in the minds of such a course.

    I have the idea that it is better to demonstrate the claim at the forefront on this turf, before evol people try to make the claim & twist up minds.

    JGuy

  6. Correction:
    “relevence” = “relevant”

  7. Sal,
    A tangent//I wrote Walter a thought that seems somewhat related to the back up system concepts you described, I think. And I’m waiting to see if he has any interesting counter thoughts. So, I completely appreciate the engineering concept here.

    I’m not certain how to best put it in words, but what I thought was along these lines (It’s more subtle, so, not as fantastik as a pure backup system as you describe)…

    Almost ALL of us humans can operate much much more efficiently than we do on a day to day basis. ie. If we eat better, excerise.. and other factors. Someone might attribute this to technology making us lazy..but I feel that is not entirely the case. Take this for granted for the moment.

    Now, if macroevolution were true, then how did that unexploited optimized system form. If it is not exploited on a daily basis, then whence did it come? How did NS select for it if it is nto being used for survival?

    Even though, your analogy with the spare tire is a more dramatic example of this concept, please consider to give thoughts on this anyone.

    JGuy

  8. I don’t understand this. No-one disputes human design (my neighbour just had a heart bypass operation). But what does it tell us, one way or the other, about design in nature?

  9. You all should check out MIT’s intro to biology. Especially the lectures by Eric Lander. Lander is a fantastic lecturer, but it is also amusing to listen to. Take any of his lectures, and compare the justifications he gives (and makes his students give) for any theory of biological phenomena, compared to the complete lack of justification for ant evolutionary statement. It’s very instructive.

  10. Its only a matter of time before patents are awarded to MIT bioengineers. If at some point, a patent is disputed by someone who claims an innovation is actually the result of random mutations and natural selection and not the work product of an MIT bioengineer holding a patent, will efforts to defend this engineer’s patent through design detection be dismissed as “psuedo-science” by a judge?

    Will this same judge be lauded by the press? Will he or she be awarded honorary degrees from various universities? Will MIT be one of those universities?

  11. geoff

    we should notice that “make random changes and see what works” is never offeredRandom mutation as an engineering solution is a non-starter. There is a reason for that.

    It’s not a non-starter. In engineering it’s called trial and error and it’s ubiquitous. Darwinians don’t like to see RM+NS called trial and error because that reveals exactly how simplistic it is. The reason there is no formal study of it is because it takes no more than a few seconds to understand and every child starts doing it instinctively shortly after birth if not sooner. In formal training in troubleshooting it’s called parts substitution – not sure what bit of something is broken? just start replacing parts until the problem goes away. Not much training required to understand that. All the training goes into how to narrow down which parts are the most likely to cause any specific problem.

  12. duncan

    No-one disputes human design (my neighbour just had a heart bypass operation). But what does it tell us, one way or the other, about design in nature?

    It tells us that the universe contains, in at least one instance, intelligent agency capable of purposeful tinkering with heritable traits in living things in steps of any size or complexity. It’s a proof of concept for ID. An equivalent proof of concept for chance & necessity does not exist except for very tiny changes of limited size & complexity. It’s sheer speculation beyond that.

  13. Good point, DaveScot. Jonathan Wells and I make the same point regarding the origin of life in our book THE DESIGN OF LIFE, which is coming out in two months:

    “Proof-of-concept works only when one proves the concept. Origin-of-life researchers are a long way from establishing proof of concept. Indeed, it has completely eluded them. Their willingness to embrace just about any highly speculative scenario for life’s origin suggests that in fact they are giving up on proof of concept and acting out of desperation, trying to shore up a materialistic explanation of life’s origin when life is clearly telling us that its origin is not materialistic.”

  14. johnnyb,

    I believe there is a setting that allows you to download the MIT video lecture to your harddrive. I looked at a couple of these last year.

    Berekely has a lot more courses which can also be viewed but they cannot be downloaded in video format. Most of Berkeley’s courses are also available in mp3 and you can download the classnotes during the semester the course is being given but since it is only audio, you have to imagine what is on the black board.

    One of the things that hit me is that here is Berkeley and MIT still using black boards while by daughter is a 5th grade teacher using smart boards in her local public school.

  15. DaveScot

    OK, so you’re saying we know external human designers can alter biology, so why not non-human, yes?

    How would you tell which was responsible? (this is Russ’s question, too)

    Does this hold outside of nature? EG: human designers can produce watches, so why not non-human?

  16. jerry -

    Also note that there is now iTunes U – you can download thousands of lectures through itunes for free.

  17. duncan

    How would you tell which was responsible?

    Without sufficient information you can’t tell. All you can infer from design is the minimal capabilities required to execute it. If you examine an old watch you know its designer had the capability to mill metal to fine tolerances. You can’t reliably infer anything more about the designer than that. When we examine the design of living systems all we can infer about the designer is some expertise in biochemistry and genetic engineering. How many instantiations of intelligent agency in the causally connected universe have or ever had those capabilities? The answer is at least one and that one instance constitutes a proof of concept.

  18. Sal wrote: “Darwinism will fail to detect and identify such designs, but the explantory filter (which all engineers use implicity) will be far more effective at identifying such designs.”
    Please describe an instance where you have applied the explanatory filter in engineering work to identify a design (or to address any other engineering issue.) Anyone?

    [Not that anyone uses Darwinism for this either.]

  19. Random mutation as an engineering solution is a non-starter. There is a reason for that.

    You’re quite wrong. Genetic algorithms are a popular technique for doing engineering design and optimization.

    Darwinians don’t like to see RM+NS called trial and error because that reveals exactly how simplistic it is.

    That is also wrong, “Darwinians” are perfectly happy to accept RM+NS as trial and error (loosely speaking — there’s no “error”, just differential survival and reproduction).

    For instance, here is arch-Darwinian Richard Dawkins:

    “A process of trial and error, completely unplanned and on the massive scale of natural selection, can be expected to be clumsy, wasteful, and blundering”.

    (A Devil’s Chaplain, pg 8).

  20. There is a hour long talk by the guy who is head of the bioengineering program at MIT. He looks about 30 years old. It is sort of an off the cuff talk and I am about 1/3 through. It is available in audio and video from ITunes U and can be watched on your video IPod or on your computer.

    ITunes is free for both the Mac and PC so anyone can access it.

  21. Sal wrote: “Darwinism will fail to detect and identify such designs, ut the explantory filter (which all engineers use implicity) will be far more effective at identifying such designs.”
    Please describe an instance where you have applied the explanatory filter in engineering work to identify a design (or to address any other engineering issue.) Anyone?

    Freelurker, good question. Do you have Bill’s book, Design Inference?

    Here is a description from that book:

    The Explanatory Filter faithfully represents our ordinary practice of sorting through things we alternately attribute to law, chance, or design. In particular, the filter describes

    how copyright and patent offices identify theft of intellectual property
    ….
    Entire industries would be dead in the water without the Explanatory Filter. Much is riding on it. Using the filter, our courts have sent people to the electric chair.

    It is clear that the Explanatory Filter (EF) is part of ordinary everyday life. It’s essential features have only been formalized in ID literature. Notice, I used the phrase “engineers use implicity“.

    At it’s root is pattern matching when there is a large space of possible alternatives — it is finding a highly improbable, but recognizable pattern. The notion of detecting patent and copyright infringement is a good illustration.

    In the practice of reverse engineering, there is a lot of mental pattern matching and analysis.

  22. Freelurker

    Please describe an instance where you have applied the explanatory filter in engineering work to identify a design (or to address any other engineering issue.) Anyone?

    You should read his statement more carefully before you bludgeon him. He said engineers use it implicity to identify designs. And further, he did not actually say the filter was used FOR doing engineering work or issues. Though, you should consider the application of reverse engineering in engineering work.

    As ReMine would say, don’t misrepresenting the arguments.

  23. Mtraven:

    Random mutation as an engineering solution is a non-starter. There is a reason for that.

    You’re quite wrong. Genetic algorithms are a popular technique for doing engineering design and optimization.

    He said it was a “non-starter”. And surely you wouldn’t suggest that genetic algorithms (GA) begin th eprocess of aiding the discovery of solutions by purely beginning with a random mutation? Isn’t there initial intelligent reasoning as the basis for what the GA is being applied? and Isn’t there an intelligent means where-by the algorithm is initialized & configured?

    At the start… a random mutation could be to reprogram & corrupt the genetic algorithm itself.

    You might argue that I misrepresetned your argument, but my argument is only to point out the nuance of what is a starter versus non starter as it was originally meant.

  24. ..clarification: The portion in the above comment that says:
    “You’re quite wrong. Genetic algorithms are a popular technique for doing engineering design and optimization.”
    …was the response from Mtraven to the block quote above it.

  25. JGuy,

    He said engineers use it implicity to identify designs.

    Yes, and I asked him to describe an instance of this. Shouldn’t be hard; he says we all do it.

    And further, he did not actually say the filter was used FOR doing engineering work or issues.
    Well, the discussion is about the activities of MIT’s Department of Biological Engineering. Sal says that use of the EF would be more effective there than Darwinism would be. Adding to this his parenthetical comment that all engineers use the EF, it is hard to escape the conclusion that he is talking about engineering use of the EF.

    While he’s clarifying that, perhaps he will also clarify how MIT would use a tool that is intended to detect intelligent causation, the EF, for the purpose of identifying how biological systems operate (their “designs.”)

    Though, you should consider the application of reverse engineering in engineering work.

    Actually, I have. Please see comment #28 here.

    In case that link does not work, here is a URL:
    http://tinyurl.com/32appj
    [I've had trouble in the past making links here. Is there a special format?]

  26. Freelurker:

    JGuy,

    He said engineers use it implicity to identify designs.

    Yes, and I asked him to describe an instance of this. Shouldn’t be hard; he says we all do it.

    I think we can all agree that it is common sensical to state that engineers know designs when they see them. And this is implicit use of the EF.

    See, implicitly, in that context does not mean the engineers open up Dembski’s book with the EF flow chart. The contrary is true. They do it instinctively. Especially when noting the experience engineers have in the design process!

    Well, the discussion is about the activities of MIT’s Department of Biological Engineering. Sal says that use of the EF would be more effective there than Darwinism would be.

    It doesn’t matter what the grand discussion is about. It matters what exact context his words were in. For example, we aren’t discussing the nature of MIT’s new course..are we? Anyway. He didn’t say what you misphrased. He said (regarding backup systems designs in life):
    “Darwinism will fail to detect and identify such designs, but the explantory filter (which all engineers use implicity) will be far more effective at identifying such designs.”

    And he’s correct in his example. Because, I ask, why would Darwinism see something as a potential back-up systems, when we know that such systems are purely foreward looking???? An engineer, presented the diagrams, would recognize this functional design much faster than a Darwinist with the expectation of the something akin to an ad hoc nature.

  27. To help clarify the point on the use of the word implicit… where I wrote about engineers detecting designs, I said:
    “See, implicitly, in that context does not mean the engineers open up Dembski’s book with the EF flow chart. The contrary is true. They do it instinctively.”

    I will append the above with this:
    “While Dembski’s has, simply speaking, outlined that innate process as the Explanatory Filter.”

  28. JGuy,

    If you had attempted to present an example of an engineer using the EF (implicitly or otherwise) you would have discovered that you, like Sal, are equivocating on the word “design.”
    The EF is intended only to detect intelligent causation. It is not intended to identify the structure of a biological system or to identify how a biological system operates. For example, it will not identify the use of back-up systems. [Darwinism, of course, will not either; that wouldn’t make any more sense than using the EF.]

  29. 1st posting for long-time lurker.

    As a structural engineer, this thread is fascinating to me. In a layman’s attempt to continue Freelurker’s line of thinking, I believe the EF is more of an analysis tool which enables the detection for the effects of design. I would not think the design process itself would be subject to EF. I have never once thought of the finished structure as a product of chance and necessity (C+N) – the first 2 filters of EF – since the two are so far removed from the final product (i.e. iron ore and various elements which form in the crust. While their formation can be considered a product of C+N, the elements go through so much intentional manipulation to create rolled steel shapes that C+N no longer plays a significant role in the end).

    Slightly off topic, I’ve noticed redundancy is used as proof for NDE. However, in engineering, redundancy plays a major role in the design process. From a structural viewpoint, it is extremely important that the various loads applied to the structure have multiple paths to the ground. Another way of putting it is redundancy ensures that the failure of a single member will not cause a progressive collapse of the entire structure. An example of this is the Confederation Bridge in Canada.

    Therefore, if redundancy plays a significant role in the design of buildings, bridges, and other engineered items, it can also play an important role in the design of nature. Looks to me like the redundancy argument can be made equally in favour of both design and NDE, kind of like homologies.

  30. Freelurker,
    I agree with much of what you said. Maybe even the equivocation part – though it would not be deliberate. Keep in mind that a literal design inherently has an intelligent cause. And ‘design detection’, as far as I understand, is equal to the term ‘detecting an intelligent cause’… [not detecting what function does the design(s) provide or perform?] That part would be an onus on the engineer’s raw expertise & intuition. And if he/she is not biased to think there is no intelligent design present, then he/she will not be as tentative relating the system to itself or the whole. IMO – this will lead to more & faster discoveries.

    Again, I can see where you might think design is being equivocated; however, nobody here is saying the EF will tell you that this orderly clump of matter(design) is a system designed for doing XYZ or QRS. But you can, I’d argue in most cases, still determine if the clump is designed using the EF (implicity).
    Example: I can be handed some device or gadget, and not know what it’s function or purpose is, yet I can know it is designed.
    Then by simply unbiasedly recognizing that there is a design present, one will then inherently be inclined to make a successful discovery of an actual succinct description/abstraction of that design [eg. in naming it's function and/or purpose]. I don’t think that the engineer employs EF to do that part specifically; however, EF is used in the process to identify the trail intelligent activity has left(akin to following fingerprints).

    JGuy

  31. Reading some threads the other day about how much information might be stored in DNA and how, I got to thinking that ID might make predictions about how studying DNA could lead to break-throughs in information systems such as the following:

    - New methods for writing error correcting code (ECC)
    - New software engineering design patterns
    - New compression techniques
    - New methods for storage, distributed processing, clustering, or other forms of data and process redundancy
    - New approaches to parallel processing

    Does anyone know if these kinds of predictions have been made in a more formal manner?

  32. JGuy,

    “… nobody here is saying the EF will tell you that this orderly clump of matter(design) is a system designed for doing XYZ or QRS…”

    Well, there is still the matter of the EF identifying backup systems. Sal has not changed or clarified his position.

    “But you can, I’d argue in most cases, still determine if the clump is designed using the EF (implicity).”

    And I would argue otherwise, but I don’t have some new argument to introduce here. (My favorite critique is “The advantages of theft over toil: the design inference and arguing from ignorance.” http://tinyurl.com/hqu4l)

    “Then by simply unbiasedly recognizing that there is a design present, one will then inherently be inclined to make a successful discovery of an actual succinct description/abstraction of that design [eg. in naming it’s function and/or purpose].”

    Here again, the semantics can be treacherous. This is the area that MIT correctly refers to above as engineering analysis; it’s the part where they “advance fundamental understanding of how biological systems operate.” When people perform engineering analysis they produce models. Saying that they are discovering designs is only adding unnecessary and unsupported teleological assumptions.

    The upshot of all my previous comments is that the work of MIT’s Department of Biological Engineering does not fall readily under either ID or Darwinism; it belongs under Engineering.

  33. Sal,
    Your comment dated 07/05/2007 (currently #21) did not appear until some time yesterday, 07/08/2007. It must have gotten stuck in a queue. I guess it can happen to anybody.

    ”Freelurker, good question.”

    Actually, I didn’t ask a question. I asked you to provide an example that would back up a claim you made.

    You claimed that the EF could be used to identify designs such as backup systems. Please show how the EF, a tool that is intended to do nothing more than attribute things to law, chance, or design, can do this.

    You don’t have to go through the steps of the EF; just start from the end of the EF process, where you have already determined whether or not there was intelligent causation. Then take us to where you have identified a backup system.

    From here, it looks like you are just equivocating on the word “design.”

Leave a Reply