Imagine, if you will, that you are present in the Garden of Eden, a background character at the big show. To your left is an important tree with some apples on it, but there are also other trees in the same orchard. If the apple tree will give you knowledge, imagine that the pear, plum, peach, and lemon trees all have something to offer, too: let’s say, culture, economics, politics, and infrastructure. Whatever you decide to try is the foundation of society that you can take with you outside of the garden. Don’t eat the apple: we know what happens if you take a bite out of that. Which would you choose? … [please read below the rest of the article].
Campos, Leo, Kendall Giles, Tiffany Smith. 2021. “Tributaries on a Heraclitian River: A Collaborative Review of Joseph Pitt’s Heraclitus Redux.” Social Epistemology Review and Reply Collective 10 (5): 6-13. https://wp.me/p1Bfg0-5OA.
🔹 The PDF of the article gives specific page numbers.
❦ Aberdein, Andrew. 2020. “Review of Joseph C. Pitt, Heraclitus Redux: Technological Infrastructures and Scientific Change.” Social Epistemology Review and Reply Collective 9 (7): 18-22. https://wp.me/p1Bfg0-5cJ.
Heraclitus Redux: Technological Infrastructures and Scientific Change
Joseph C. Pitt
Rowman & Littlefield, 2019
Joseph C. Pitt makes a compelling case for technological infrastructure as the enablement that humans have selected to make sense of our complicated world. While likely the least glamorous of the fruit trees described above, infrastructure encompasses and supports all of them in one way or another. In Heraclitus Redux: Technological Infrastructures and Scientific Change, the resources, humans, and structures that the environment encompasses are essential and formative to scientific research. In fact, scientific research is not possible without them.
My goal in this review is to show how Pitt’s argument works in the context of infrastructure studies, which seeks to identify the remarkable power of the technology, organizations, and conceptual categories people use to demarcate their worlds and demonstrate how that infrastructure guides decisions, actions, and outcomes. For more on abstract knowledge and the concept of logos, please see the parallel essay by Campos, which addresses important considerations about scientific progress and how we conceive of knowledge as “real”. For more on technology’s relationship to scientific change and its results for humans and society, please see the essay by Giles, which places Pitt’s assessments and conclusions in the context of technology-enabled humanity and ponders how we might change course.
Pitt argues that, because scientific knowledge can only be obtained through technological infrastructure, our knowledge is limited to that which we can gather through whatever infrastructure we have at present. Our ability to grasp knowledge is forever overwritten as our capabilities develop and change (enter Heraclitus, who noted the impossibility of entering the same river twice). As a result, Pitt argues, the concept of scientific progress is a delusion, because we can have no evidence of advancement, simply evidence of a changing infrastructure with varied developments and a hopefully more cohesive story.
There are consequences to relying on a technological infrastructure, Pitt reminds us. The laboratories we build to explore and understand a certain problem or the grant funding we provide to support this exploration make the researchers beholden to the task defined at the outset of the relationship. The information infrastructures we create to support knowledge flow can also stymy our capacity to share knowledge. We create metaphorical “crosswalks” to make sense of data across disciplines, and then we create crosswalks among the crosswalks to ensure traceability to the origin. We do not, however, easily acknowledge the unfathomable nature of the knowledge we are trying to hold within contexts that are not equipped to contain it.
“The knowledge creation process is always ongoing, and the results are always in flux,” Pitt notes in the context of his argument for a pragmatic theory of knowledge, where “truth is always tentative, contingent upon being a guide to successful action” (44). Allowing for tentative truth gives the researcher the opportunity to use and apply that tentative knowledge to different aims. We do not discount the knowledge that we think we have obtained because it is by necessity imperfect, but we accept it and we use it to deliver other, related knowledge. Over time, we may lose the traceability to its origins.
In addition, Pitt tells us, knowledge is always dispositioned in community. Peer reviews enable entrance into the society of knowledge creators. Whether or not knowledge belongs in one discipline or another is negotiated by funding organizations, scientists, academic departments, and publishers. No one individual is the final arbiter of any item of knowledge. The technological infrastructure that creates knowledge is the same infrastructure that decides its position, relationships, and value.
If a scientist is persuaded by this argument for infrastructure as guiding force, how might it be used to his or her advantage? Well, the scientist would work the system. The scientist would decide that she doesn’t especially need to be “Right”, she just needs to be able to contribute. She would participate and discover and publish or teach, and she would get to do so again and again over the course of a career, and at least once her contribution would hopefully be a little bit closer to an unfathomable truth, rather than a little farther away.
In this concise book, Pitt gives the scientist permission to participate in the technological infrastructure. He helps the scientist to see that the technological infrastructure that equips us to learn and develop knowledge can also constrain us. By acknowledging and assessing the consequences of technological infrastructure, the scientist can consider which options he or she is positioned to accept or not, and he or she can then decide whether and how to address the unseen infrastructure that may shape scientific research results.
Taking on the robes of the Heraclitian is no easy endeavor, Pitt tells us. The challenge the Heraclitian endures is to remain aware that any knowledge you develop or contribute or provoke will ultimately be proven wrong by a future generation, and yet you take on the knowledge task anyway for the very sake of that future generation (59). Accepting Pitt’s argument brings with it a dose of humility and a responsibility to contribute.
Pitt notes in conclusion, “I believe that the concept of a technological infrastructure gives us a powerful tool to think about how we live and to change the ways we live, when needed” (106). One might accept that we will never truly enjoy the fruit of the apple tree, but a pear could be delicious.
I would like to focus on one part of the Heraclitian “river” that Pitt brings up in his argument: progress in science and technology. But before going tackling the issue of progress it is important to note that a comparison between Heraclitus and Plato is not fair. A much better foil would be Parmenides, as these two pre-Socratics and their radically opposing views were in a sense what Plato had to wrestle with and resolve.
One of the great things about the pre-Socratic philosophers is that they were the first (in Greece anyway) to notice and think about the predictable and consistent laws of nature and to explain their insights into these workings as something which is understandable by humans using reason. As the first natural philosophers they began an exercise of self-conscious inquiry into nature without the reliance on myths. Through their efforts they were able to conceive of abstract knowledge and to offer universal claims about how things work without resorting to supernatural explanations, or at least not mythological ones.
What I would like to think through is how the concept of logos could help with the thorny issue of progress—and I would recommend also reading the parallel essay by Smith on how Pitt tackles technological infrastructure and the difficult problem that just because we have “evidence of a changing infrastructure with varied developments and a hopefully more cohesive story.”
While the classical, and certainly the pre-Socratic concept of logos is a slippery one to pin down, there is no doubt that Heraclitus felt it was important. Mostly, I believe, because “only change” is a chaotic affair, and one could assume, to the empiricist with an open mind the very sensory experience being called to give evidence of the logic of the arguments demonstrated a form of order, a coherence or harmony. How to account for that if everything is only change? With a concept such as Heraclitean logos we can see that change is bounded by some order which gives it shape. With this lens one can look at technological or scientific change as not willy-nilly but also not as completely determined. Perhaps the logos I propose is a form of soft determinism, though I think it is a broader concept than that. In my view determinism falls within the logos. In a Heraclitean universe, underlying all the change it is organized change, based on logos—all things change but the logos remains. In this case, while the river is never the same, there is still The River and my Self to step in it, even if both of those are changing.
Do we need progress to have science? The scientific process, as Pitt points out, is one of trial and error, but can we see scientific error as progress? Pitt does not seem to think so. But “error” in the case of scientific experimentation is a form of progress. To know that a thing A is not “it” (whatever the goal of the experiment), while it still leaves me with a lot of other things which maybe “it”, I can at least discard A. That’s progress, no? In fact, it is through discarding things which you progress. That is what trial and error mean, it is not “trial and succeed”.
Pitt begins his book by pointing out that science is a set of multifaceted social processes in constant state of change. Further, there is no “one science” with many sub-disciplines, which is the traditional view of science. In fact, there are many methods employed by the various sciences with little in common with each other. One thing that is not made explicit but is worth raising here is that one of the factors that differentiates sciences is the apparatus. Might it be possible to create a new taxonomy of science by classifying the types of equipment that is required? In this building we conduct microscopic sciences, while in that building over there, we conduct sciences with Bunsen burners and test tubes. Over there in that building with golden domes and the parking lot full of Mercedes is where we conduct Big Science.
The importance of this technological distinction is what we can discard when trying to understand science and technology. There is no need to fixate on universal answers, or a Theory of Everything, or perennial philosophy. But this is where the lack of a “logos” leaves us rudderless or having to explain everything through a social constructivist lens. As Pitt points out when talking about Peirce: “the real or reality is independent of what anyone thinks. That is different from assuming that what we say is real is, in fact, real. We can investigate all we want and come up with any number of theories, but that does not change what actually is out there, even while what is actually out there is changing” (68).
Pitt leans heavily on the word “progress”. The way it tends to be used when talking about science almost makes one think of the mythological gods which Heraclitus was questioning. As Pitt points out “There is no progress. Science changes but progress implies something else” (67). Indeed it does. If I am going to measure the progress of something trivial, such as how close am I to my destination, I need to know what that destination is. But if we cannot accept, or even if we accept, we cannot define what the destination of science is, then how to know we are making any progress at all. Many examples are raised in the book to question whether there has been any advance. As Pitt points out: “If there is no coherent account of a goal for scientific progress, it would then seem like we also can’t measure that progress, for that would be to determine how much closer to the goal we have come” (69).
Pitt is reluctant to impose an a priori direction to science and technology, but perhaps we could talk about order? This avoids the pitfalls of defining a destination which is untenable these days, while at the same time allowing for something which embraces both the social forces within which science occurs as well as the fundamental unquantifiable, by which I mean breakthroughs which happen despite society rather than because of it, which science has been able to produce.
Pitt focuses the importance of what he calls the “fundamental law of technological development: unintended consequences” (80, italics in original) and goes on to describe that the key issue is not technological progress or technological determinacy but how to assess technological development: “Assessing how they impact society, and the environment cannot be accomplished successfully without including how they impact each other and what the unintended consequences of all those interactions might be” (81).
Just because it is nearly impossible to accurately, or even usefully, model the impact of a new technology does it mean we should not try? In the absence of alternative risk assessments (see at a minimum Shelia Jasanoff’s thoughts on this) the current default ones will remain in use—yielding roughly the same results.
To circle back to Heraclitus’ original antagonist: Parmenides. Among his claims to fame were that there is no such thing as movement. The argument goes:
P1: What is not cannot exist.
P2: Things cannot change—because a change implies a move from what it was into what it is not (any longer). So, a change is a move from What is Not -> to what it is now.
P3: Movement cannot exist! Because movement implies that you are no longer in a place you were, that old place does not exist.
Therefore: Whatever exists in now must be as it always was.
Conclusion: The universe is uniform (unchanging) throughout. It is never what it is not.
Of critical importance for his argument is his first premise: What is not cannot exist. Let us now flip this into a more positive form, what I am calling the logos, and say that “For this (technology/event/thing) to exist all other things which are affected by it must also exist.” For a telephone to exist all sorts of physical, technological, social, economic, historic things must also exist together. If they do not exist, then neither does the telephone.
I think the logos concept goes a long way in dealing with questions as to why certain advancements happen at some places and not others when at least the basic components were already known. Logos, of course, is not the same as telos, something may be comprehensible without the necessary end point being in sight. In this it also serves to protect us from some sort of technological determinism. In fact, if you accept Pitt’s argument around the unintended consequences of technological development logos might be your last best hope towards developing a sufficiently robust narrative to curtail rampant development without getting enmeshed in deterministic arguments.
Evidently Joseph Pitt uses a different editor than most because the title of his latest work, Heraclitus Redux: Technological Infrastructures and Scientific Change, contains neither the words “innovation” nor “AI.” Innovation and AI are such current concerns in society that they have even subsumed the staid halls of higher education—Yale Law School now has an “Entrepreneurship and Innovation Clinic” and the University of Florida has announced plans to incorporate AI concepts, technologies, and methodologies into the entire university’s curriculum and to make AI a defining element of the university’s brand. Even my own university, Virginia Tech, is building a completely new campus from scratch, officially called the Innovation Campus, in our nation’s capital region. But do not get me wrong—erudite discussions of technological innovation and the powers of AI to change even science itself are contained inside Pitt’s short but thoughtful book. Pitt however focuses on something I think even more profound than current technology trends. As a society, rather than just rewarding the rush to crank out ever more gadgets that, upon critical analysis, are neither innovative nor intelligent, in a time when our technologies are becoming so complex and their impacts on society are potentially more devastating Pitt is suggesting that it is in our best interests to understand the pervasiveness, power, and potential of technological infrastructures in helping us build a better future.
This book is a succinct, clear, and coherent discussion of scientific change and its relationship to technological infrastructure, though sometimes the chapters seem too brief. An update to earlier accounts of scientific change that tended to minimize the essential role technology has played in science, Pitt’s thesis is that technology not only makes the endless frontier of science possible but is the fuel itself. Early chapters establish that far from a pure field called Science, there are actually multiple independent sciences, and in each science the steps and standards for conducting that science have evolved over time. Galileo’s telescope,
for example, changed not only how science was conducted but also an entire worldview. Similarly, the electron microscope is shown not only to change what counts as an observation, but also to demonstrate science’s reliance on technology to better understand the world around us.
The telescope and the electron microscope are just two elements of what Pitt describes as technological infrastructure “the range of things that make science possible: funding agencies, universities, private corporations, technicians, labs, graduate students, journals, and so on” (34). For example, all the technologies and instruments that make up the different science labs are what makes science possible, and since it is engineering research that makes these technologies and instruments, Pitt suggests that engineering research is even more fundamental than science research. And while technological infrastructure may not have always been such an important part of early science, it is absolutely essential for modern science. Moreover, the larger and more complex this infrastructure becomes—think about the massive in size as well as cost KATRIN (Karlsruhe Tritium Neutrino Experiment) project to study the infinitesimal neutrino—the more science depends on this infrastructure to shape its theories, to interpret its results, and to determine theory success or failure.
With this foundation of technology as the key to scientific change, Pitt focuses on the elements and importance of this technological infrastructure in Chapters 8-10. Pitt realizes that his earlier (Thinking About Technology, 1999) definition of Technology as “humanity at work” (77) did not quite hit the mark, especially considering that today’s technologies such as automation threaten at a minimum to change the work done by humanity, while some even despair of future unemployment at the hands of algorithms and robots. But in that realization is the core of what Pitt says is wrong with society’s drive to produce more and more gadgets, to develop more and more code, to embed technology into every nook and cranny of our physical world—what Pitt calls the “fundamental law of technological development: unintended consequences” (80).
The problem is that not only are the technologies fueling scientific change ever larger and more complex, but they are also being released at what seems to be an accelerating pace into society—a cycle of repeated development and production all without sufficient consideration, oversight, and due diligence. These technologies are so complex and so powerful that we don’t know how they will affect the world. Steer an oversized cargo ship into the side of a canal—whether due to wind or human error, it does not matter—and in an instant billions of dollars in world trade grinds to a halt. Snake a 330-mile-long natural gas pipeline from the Marcellus Shale in West Virginia over hill and dale across Virginia, an example described by Pitt, just to satisfy the energy needs of a growing consumer culture, and the potential impacts on the lives and local economies that must deal with the results are even greater. Next, embed AI, sensors, and automation into coffee pots, cars, trucks, cargo ships, stock markets, robots, surveillance cameras, and hospital equipment, while networking them all together without sufficient or even any security mechanisms, and perhaps you can understand Pitt’s concern about unintended consequences.
Just as the telescope changed not only how science was conducted but also our entire view and understanding of the world we live in, so too does technology ask us to reconsider our understanding of ourselves. Pitt reworks his definition of Technology to be “technological infrastructures make humanity at work possible” (102), incorporating ‘technological infrastructures’ for ‘technology’ and adding ‘possible’ to stress our own agency and potential in a world that we share with technology. While that helps his original definition of Technology, I am not sure the new version goes far enough because of the preposition ‘at work,’ not to mention what we mean by ‘humanity.’ I mentioned automation earlier and this new definition still does not sufficiently resolve that bugbear. With the increasing capabilities of algorithms and machinery to automate tasks, sense the world, and make decisions independent of human control, we have the chance to change the nature of the world we live in, what roles and responsibilities we have when we “work,” how much leisure time we have, and how we want to construct meaningful lives. Pitt describes robots at a company called Zymergen that literally design and perform science experiments in the lab, experiments that the human scientists there did not create nor possibly even understand.
We are witnessing another change in how science is done, for “if science is what scientists do, and if more and more of the ‘science’ is being done by machine learning systems, then is it really science?” (86). More importantly, should we write “The scientists who performed the experiment” or “The scientists that performed the experiment?” Do we refer to the robot designing and performing the experiment as a “who” or a “that?” Who or what is the “taxi driver” when the human “driver” sits in the front seat, perhaps watching a movie or playing video games and not actually driving at all, while the self-driving car drives everyone from point A to B? When our lives and bodies are increasingly merging with technology, what exactly do we mean by “humanity?” Pitt says “for the pragmatist, the mark of knowledge is successful action” (6). That definition sounds very similar to a description of AI, which is all about making successful actions in an environment towards some goal.
As we develop ever more complex and powerful technologies, yes our science becomes more complex but so too do our societies. “Who” vs “that” distinctions are blurring. There is momentum carrying us forward, but just as the river is different each time we visit it, Pitt’s Heraclitian philosophy of technology suggests that we have the potential to change course—our future is not determined, but only if we understand the importance and power of technological infrastructures and their role in helping us live better lives and build better futures. Pitt’s framing suggests a promising starting point for further explorations of the relationships between science, technology, and society, and it is clear that humanity needs to get to work.
Categories: Books and Book Reviews