Abstract
There are choices about the number of ways to approach and understand a problem. Sometimes finding the one right analytical approach is sufficient. Other times, such as with the Manhattan Project, the use of many approaches is desirable. Increasing independence among multiple analytical approaches, i.e. using a pluralistic approach, can be a good strategy to get knowledge to make decisions and understand a system. We considered two frameworks that have attempted to provide advice to engineering and scientific practitioners on when and how to use multiple analytical approaches. The RAND Corporation’s parallel path strategy, as described by R.R. Nelson, is a way of using independent engineering efforts to explore what parts of the design space are feasible, as well as what the cost and schedule would be for different designs. Richard Levins and William Wimsatt’s focus on model independence provides motivation and insights for using multiple models to assess the same system. While these approaches may appear different, both rely on using a group of analytical approaches where the individual members are independent – or different from—one another. Comparing these two approaches provides suggestions about how to utilize independence to address uncertainties in design and model-systems. We argue that the deliberate creation of independence among engineering developments and models should be tied to key uncertainties in the model or system. With relatively low uncertainty, choosing one approach may be acceptable. Both suggest that there can be (but are not always) benefits from using multiple approaches, which can increase accuracy and reduce cost. Using a few independent approaches – as opposed to many – may be more desirable when there are only a few bounded uncertainties about the system.
“No one will deny that a problem cannot be fully formulated until it is well on its way to solution. The real difficulty, the nub of a problem lies somewhere amongst the subproblems…The nature of the problem can only be found by examining it through proposed solutions and it seems likely that its examination through one, and only one, proposal gives a very biased view. It seems probable that at least two radically different solutions need to be attempted in order to get, through comparison of subproblems, a clear picture of the ‘real nature’ of the problem” (Marples 1961, “The Decisions of Engineering Design. p. 64, source found due to Lenfle 2011).
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
The views expressed here are the authors’ and do not necessarily represent the views of NASA or the United States Government.
- 2.
Levins evaluated models based on the realism, generality, and precision and claimed there was a tradeoff between these three features, as discussed later. This led to pursuing three types of models for his primary analysis case in the ‘Strategy’, but he did not have a general rule calling for three different models. However, he was concerned with the manageability of models, and the use of 2–4 models may be at the upper end of what’s manageable. He recognized there could be other dimensions for evaluating models. For Nelson, different assumptions could yield different results on the desired number of paths. Also, we note that while Levins and others use the term ‘robustness’ to refer to this process of agreement across multiple lines of evidence, we focus on the term independence because it is less commonly used and is less likely to be misconstrued (Lempert et al. 2006).
- 3.
As has been explored by Lenfle and Loch, the 1950s RAND trandition of espousing parallel paths was in part pushed aside for cultural reasons for much of the last 50 years. Lenfle convincingly argues that this RAND literature stopped being cited in the 1960s due to U.S. Secretary of Defense Robert McNamara’s desire for a more stream-lined management approach (Lenfle and Loch 2010). This McNamara-led approach led to the creation of the ‘rational’ Stage Gate model of innovation, which assumes that innovation proceeds in a step-by-step fashion and that engineering managers should solely pursue development of a single effort at a time in order to keep costs small. While this stage gate linear model has been much discredited by historians and scholars of innovation, no competing theory of innovation has successfully replaced it (Godin 2006; Szajnfarber 2011).
- 4.
The two main caveats are about certainty of the design and the role that schedule pressure plays in motivating his argument. First, the argument is not a certain one, and it is always possible that another approach could be successful. For example, he said “It is true that the atomic-bomb project, the method that actually produced chain-fissionable material for the first bombs was considered relatively unpromising early in the program…But it might well be argued that had all our money and effort been allocated to this latter method, it might have produced material not only just sooner, but sooner than the former method” (p. 362). However, Nelson’s argument and calculations suggest that the use of parallel approaches will be right more often than not. Second, Nelson indicated that time concerns may be the biggest reason to embrace a parallel paths approach (p. 361): if one does not care about time, then you can wait for more research to help provide clarity on what development option should be pursued.
- 5.
Since that time, the climate science community has written a significant amount of literature on that topic, with the lead scholar of independence analysis, Knutti (2018), offering up much more detailed approaches to align different climate models in deliberately independent ways. There are also recent examples in the engineering literature on the role of interdepednence across parts of a portfolio of technology projects (Wicht and Szajnfarber 2014).
- 6.
This feature emerges in a case study we have performed, wherein two groups tried to build the same sensor but had different pathways for what piece of superconducting metal would be used in the instrument. Sometimes the exact shape of the superconducting bilayer would differ based upon which technician was handling the software. The resolution of the sensor could differ significantly based upon the pathway taken, but there was not sufficient understanding of the phenomena to enable predicting the outcome of the physical instantiation independent of letting the development teams create the sensors. For more details on these dimensions, see Pirtle (in preparation).
- 7.
From some of our other research (Szajnfarber 2014; Pirtle, in preparation), we have studied a case where two groups tried to build the same functional device at the same time, but went about designing their system differently. For one development team, it turned out that the way of setting up the subsystems of the system, aligning electronic readout systems with the X ray detectors, was a significant barrier to progress. The other team set up the interfaces between its subsystems differently and was able to succeed, although both teams were able to exchange helpful knowledge and information across both paths.
- 8.
To what extent is knowledge really the desired outcome across these two strategies? One might say that it the resemblance is superficial: that model independence is about truth, not about optimizing designs; and the parallel paths strategy is about increasing practical odds of success of yielding an outcome from a portfolio. In this sense, it wouldn’t be clear that that there is much commonality across the two strategies, and that it may not make sense to refer to them both as employing diverse analytical approaches. There is some merit to this, as one could employ these strategies while holding different intentions. However, as described above, the articulated rationale behind the parallel path strategy is about how it helps decision makers due to the knowledge that is accrued. The parallel path approach wants to know what designs are possible and at what cost and schedule. Engineers can face major uncertainties about what exactly their final design will be, and in this sense both approaches are focused on the creation of knowledge.
References
Alchian, A. A., & Kessel, R. (1954). A proper role of systems analysis. RAND Corporation. D-2057.
Baldwin, C. Y., & Clark, K. B. (2000). Design rules: The power of modularity (Vol. 1). Cambridge, MA: MIT press.
Cetina, K. K. (2009). Epistemic cultures: How the sciences make knowledge. Cambridge, MA: Harvard University Press.
Giere, R. N. (1999). Science without laws. Chicago: University of Chicago Press.
Godin, B. (2006). The linear model of innovation the historical construction of an analytical framework. Science, Technology & Human Values, 31(6), 639–667.
Hacking, I. (1981). Do we see through a microscope? Pacific Philosophical Quarterly, 62(4), 305.
Hitch, C. J., & McKean, R. N. (1960). The economics of defense in the nuclear age. Cambridge, MA: Harvard University Press.
Hounshell, D. A. (2000). The medium is the message, or how context matters: The RAND Corporation builds an economics of innovation, 1946–1962. In Systems, experts and computers. The systems approach in management and engineering in World War II and after (pp. 255–310). Cambridge, MA: MIT Press.
Houkes, W. (2009). The nature of technological knowledge. In A. Meijers (Ed.), Handbook of philosophy of technology and engineering sciences (pp. 309–350). Dordrecht: Springer.
Klein, B., & Meckling, W. (1958). Application of operations research to development decisions. Operations Research, 6(3), 352–363.
Kroes, P. (1998). Technological explanations. Techné: Research in Philosophy and Technology., 3(3), 124–134.
Knutti, R., (2018). Climate model confirmation: From philosophy to predicting climate in the real world. In Climate modelling (pp. 325–359). Palgrave Macmillan
Lempert, R. J., et al. (2006). A general, analytic method for generating robust strategies and narrative scenarios. Management Science, 52, 514–528.
Lenfle, S. (2011). The strategy of parallel approaches in projects with unforeseeable uncertainty: The Manhattan case in retrospect. International Journal of Project Management, 29(4), 359–373.
Lenfle, S., & Loch, C. (2010). Lost roots: How project management came to emphasize control over flexibility and novelty. California Management Review, 53(1), 32–55.
Levins, R. (1966). The strategy of model building in population biology. American Scientist, 54(4), 421–431.
Levins, R. (1993). A response to Orzack and Sober: Formal analysis and the fluidity of science. The Quarterly Review of Biology, 68, 547–555.
Lloyd, E. A. (2015). Model robustness as a confirmatory virtue: The case of climate science. Studies in History and Philosophy of Science Part A, 49, 58–68.
Loch, C. H., et al. (2011). Managing the unknown: A new approach to managing high uncertainty and risk in projects. Hoboken: Wiley.
Marples, D. L. (1961). The decisions of engineering design. IRE Transactions on Engineering Management, EM-8(2), 55–71.
Mccurdy, H. E. (2001). Faster, better, cheaper: Low-cost innovation in the US space program. Baltimore: JHU Press.
McNie, E. C., et al. (2016). Improving the public value of science: A typology to inform discussion, design and implementation of research. Research Policy., 45(4), 884–895.
Nelson, R. R. (1959). The simple economics of basic scientific research. The Journal of Political Economy, 67, 297–306.
Nelson, R. R. (1961). Uncertainty, learning, and the economics of parallel research and development efforts. The Review of Economics and Statistics, 43, 351–364.
Nightingale, P. (2009). Tacit knowledge and engineering design. In Meijers, A. (Ed.), Handbook of philosophy of technology and engineering sciences (pp. 351–374). North Holland.
Odenbaugh, J. (2003). Complex systems, trade-offs, and theoretical population biology: Richard Levin’s “strategy of model building in population biology” revisited. Philosophy of Science, 70(5), 1496–1507.
Odenbaugh, J. (2006). The strategy of “the strategy of model building in population biology”. Biology and Philosophy, 21(5), 607–621.
Orzack, S. H. (2012). The philosophy of modelling or does the philosophy of biology have any use? Philosophical Transactions of the Royal Society of London B: Biological Sciences, 367, 170–180.
Orzack, S. H., & Sober, E. (1993). A critical assessment of Levins’s the strategy of model building in population biology (1966). The Quarterly Review of Biology, 68(4), 533–546.
Parker, W. S. (2011). When climate models agree: The significance of robust model predictions. Philosophy of Science, 78(4), 579–600.
Pirtle, Z. (2010). How the models of engineering tell the truth. In I. van de Poel & D. Goldberg (Eds.), Philosophy and engineering: An emerging agenda. Dordrecht: Springer.
Pirtle, Z., et al. (2010). What does it mean when climate models agree? A case for assessing independence among general circulation models. Environmental Science & Policy, 13(5), 351–361.
Pirtle, Z., et al. (2018). Engineering model independence: A pluralist strategy to encourage independence among models. Techné: Research in Philosophy and Technology., 22, 2.
Pirtle, Z. (in preparation). Engineering independence: An analysis framework for systems engineering in complex development programs. Dissertation, Department of Engineering Management and Systems Engineering, The George Washington University
Poznic, M. (2016). Modeling organs with organs on chips: Scientific representation and engineering design as modeling relations. Philosophy & Technology, 29(4), 1–15.
Scherer, F. (2011). Parallel R&D paths revisited. Cambridge, MA: Harvard University, John F. Kennedy School of Government.
Szajnfarber, Z. (2011). Innovation pathways in technology intensive government organizations: Insights from NASA. Dissertation from the Massachusetts Institute of Technology.
Szajnfarber, Z. (2014). Space science innovation: How mission sequencing interacts with technology policy. Space Policy, 30(2), 83–90.
Vincenti, W. G. (1990). What engineers know and how they know it: Analytical studies from aeronautical engineering. Baltimore: John Hopkins University.
Weisberg, M. (2006). Forty years of ‘the strategy’: Levins on model building and idealization. Biology and Philosophy, 21(5), 623–645.
Wicht, A., & Szajnfarber, Z. (2014). Portfolios of promise: A review of R&D investment techniques and how they apply to technology development in space agencies. Space Policy, 30(2), 62–74.
Wimsatt, W. C. (1981). Robustness, reliability and overdetermination. In M. B. Brewer & B. E. Collins (Eds.), Scientific inquiry and the social sciences (pp. 124–163). San Francisco: Jossey-Bass publishers.
Wimsatt, W. C. (2007). Re-engineering philosophy for limited beings: Piecewise approximations to reality. Cambridge, MA: Harvard University Press.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG, part of Springer Nature
About this chapter
Cite this chapter
Pirtle, Z., Odenbaugh, J., Szajnfarber, Z. (2018). “The One, the Few or the Many?”: Using Independence As a Strategy in Engineering Development and Modeling. In: Fritzsche, A., Oks, S. (eds) The Future of Engineering. Philosophy of Engineering and Technology, vol 31. Springer, Cham. https://doi.org/10.1007/978-3-319-91029-1_2
Download citation
DOI: https://doi.org/10.1007/978-3-319-91029-1_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-91028-4
Online ISBN: 978-3-319-91029-1
eBook Packages: Religion and PhilosophyPhilosophy and Religion (R0)