A new theory of surprise – unraveling the logic of uncertainty and knowledge

ABSTRACT

Intelligence is judged by its ability to avoid surprises. Unfolding the dry logic of uncertainty shows how intelligence failures caused by surprises are unavoidable. This is done by the systematic study of the general conditions that trigger the different kinds of surprise, categorized according to their different epistemological nature. Five conceptual categories of surprise are analyzed: lack of expectations; insufficient evidence between two or more alternative courses of action; false justified beliefs; incomplete theory to ground a belief, and an utterly false expectation produced by our (usually reliable) mindset. Intelligence serves as a countermeasure to opponents’ actions at strategic and tactical levels. However, intelligence analysis can only minimize the risk; it cannot eliminate it altogether.

Acknowledgements

I am deeply grateful to Ralf Lillbacka, Terry Quist, and Roman Kolodii for their precious suggestions and comments. I want to thank Sidharth Kaushal for a friendly discussion on the paper topics that helped me improve it. I am grateful to three reviewers for their comments that furtherly helped me in improving the paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. But only idiots are always surprised by the same event over again, see Aristotle, Metaphysics, I, 982b, 1.

2. Wirtz, “Theory of surprise”, 97–111.

3. Handel, “Master of war”, 165–92.

4. Handel, “The problem of strategic surprise”, 229.

5. The so-called surprise effect, though labelled to suggest its being a result of surprise, it is often confused to be the surprise itself. This major confusion is the origin of critical misunderstanding of the topic of surprise, including the author of this essay. However, trying to disentangle important early comments, I arrived at the conclusion that there is a general feeling that if a surprise is not followed by any surprise effect, then it is difficult to conceive it as a surprise. I will not tackle the point explicitly in this essay, but it is indeed crucial that surprise is neither defined nor evaluated (as such) from this standpoint. Surprise and the surprise effect are two different things, though clearly related by a causal relationship at a psychological level.

6. Betts, “Surprise Despite Warning”.

7. Pili, “Deciphering intelligence analysis”.

8. Lillbakca, “Schelling Traps as Drivers of Intelligence Failure”.

9. Ibid, especially 1–7.

10. Whaley, “Toward a general theory of deception”, Whaley, “Stratagem: deception and surprise in war”. For more see footnote 13.

11. For instance, Grabo, Handbook of warning intelligence, Grabo, “Strategic Warning: The Problem of Timing”.

12. Such as that proposed by Robert Jervis in his authoritative studies e.g., Jervis, Perception and misperception in international politics, Jervis, System effects: Complexity in political and social life. The reason for rejecting this approach is not in its value but in its relevance for this study. Jervis analysis applies for those inquiring on surprise and misunderstanding at policy level from a psychological point of view. In other words, he inquires for specific causal triggers under which surprise can manifest or be achieved. This study delves into concepts and conditions which are not causal in nature. On this more below.

13. Barton Whaley’s study on deception, although it is valuable, it deals with a different aspect of surprise, which involves disrupting the opponent’s capability to know. This paper, however, is not about factual causes and how surprise is achieved; that is for historians and empirical researchers (see footnote 12 and in text argument).

14. Although it is beyond saying, there is no need to assume that this angle of approach is absolutely the best. As other publications of mine attests (see Caligiuri, Pili “Italian Intelligence Studies Literature” and Margoni, Pili, “Social dominance orientation”), I endorsed the bottom-up approach usually preferred in intelligence studies elsewhere including empirical research. The method’s selection here is merely due to the different goal of the study at hand, and this footnote works as a disclaimer against potential objection to the method which is, after all, in line with the best ontological research (see the current developments in Basic Formal Ontology quite influential in the concrete taxonomical endeavor within the intelligence realm see Mandrick, Smith “Philosophical Foundations of Intelligence Collection and Analysis”).

15. Sorensen, “Epistemic Paradoxes”, 1.1.

16. According to some philosophers (see Sorensen in 1.1 and debate), this paradox raises for its being self-referential, namely that the statement refers to itself. Self-referentiality is a major cause of paradoxes such as the liar paradox and the like. I don’t think here the paradox is triggered by this condition, although an argument can be made for the opposite take. To see why it can, the statement can be thought as a proposition such as “I’m telling you: you will be surprised, and you will not be surprised because I told it to you”. Now, what I see here is that there is a semantic paradox only assuming what I have called “principle of transposition”, which will be introduced later in this piece of research, namely the possibility to think acts as propositions (statements, specifically.) In this way, the paradox arises by the fact that I’m talking about the coming surprise and my being its cause. Thus, the statement can be reframed as “I will surprise you and I will be the cause of the surprise”, which obviously obliterated one source of surprise (at a categorical level.) However, it leaves open the possibility of local or instance of surprise. As it will be argued, this kind of surprise is the first-type surprise.

17. Otherwise, it will not count as a surprise, as already proved.

18. See Sorensen “Epistemic Paradoxes”, 1.1 & and the following 1. sections. Though there is no space to argue here, an influential though not fully convincing analysis of this paradox can be found in Kripke “Two Paradoxes of Knowledge”, 27–51.

19. Here there is still the possibility of not having any attack at all from both sides, which would count as a surprise, after all. In this context, inaction is action.

20. Dahl, Intelligence and Surprise Attack, 29–33.

21. This statement would be interesting to explore in detail, as it leads to the complication of the definition of war itself. This is out of the scope here, but for a detailed analysis Pili, Filosofia della guerra but also McFate, New Rules of War.

22. Though many times is all about getting the right perception out of the information collected. For a very insightful study on the topic: Yarhi-Milo “In the eye of the beholder: How leaders and intelligence communities assess the intentions of adversaries”.

23. Ginor & Remez, The Soviet-Israeli War, 1967–1973: The USSR’s Military Intervention in the Egyptian-Israeli Conflict.

24. For an extensive analysis of the principle of transposition, Pili, “The Principle of Transposition”.

25. This is applicable to any realm of human action. When a person gives a flower to another one is trying to express kindness, care, and tenderness. However, associating kindness, care, and tenderness would be impossible if we don’t think the act as conducive of a specific meaning through a codified way of expression, which, ultimately, conveys a clear-enough message. Misinterpretations are, naturally, always possible.

26. An error due to a mistaken application of a given rule is always possible only assuming that the opposite possibility was available, otherwise the error would be simply the outcome of a given rule, which means it wouldn’t be an error at all.

27. Dr. Henry Kissinger fireside chat with Eric Schmidt | Talks at Google [17 April 2015]”, 51‘:00. https://www.youtube.com/watch?v=wooGL-OvA.

28. Kotkin, Stalin – Waiting for Hitler, Chap. 10.

29. Ivi.

30. Lillbacka, “Schelling Traps as Drivers of Intelligence Failure”, 10.

31. Zubok, A failed empire: the Soviet Union in the Cold War from Stalin to Gorbachev, 150.

32. Caesar, The Gallic Wars, I.53 - I.54 (Chapter 44, p. 190), italics added.

33. Naturally, it can be repulsive to us reading Caesar’s erudite observation on how to manage counterinsurgency operations, although nobody can say it wasn’t very effective. But even the judgment over the effectiveness can be debatable. What instead is completely beyond doubt is our ability to follow Caesar’s argument and perfectly grasping the meaning of his words. This should sound as a pretty remarkable achievement, as after more than two-thousand years we are still able to see what the reference of Caesar’s words is. If we compare our understanding of other portions of ancient literature and way of thinking, we must be all in agreement in stating that, along with mathematical statements and geometrical analysis, warfare looks a universal intelligible language, which was an argument made for chess as for struggle – see Lasker, Struggle, Pili, Scacchi e Filosofia della Guerra (Introduction and chap. 1 & 2).

34. Interestingly, all the philosophers who assumed possible to deny the possibility to access others’ minds (a particular kind of skepticism) should be challenged in this assumption by the fact that human military history (at least) wouldn’t be explainable without postulating others’ minds and humans’ ability to read it. In line with Terry Quist’s early observations about this paper, I am not arguing for total blindness toward others’ behavior. It is just a matter of the fact that others’ minds are directly inaccessible (that is why the practice of torture is so widespread and universal). For this see Pili, Filosofia pura della guerra, 269–304.

35. There is a big difference between group and singular subjects, which plays a major role in misconceptions and misunderstanding in the intelligence domain – Pili, “Be Coherent with Yourself!”, 100–101.

36. Maizland, “Why China-Taiwan Relations Are So Tense”.

37. More on this below. For an interesting related discussion of this point, see Quist, “What philosophy can do for intelligence”.

38. For a classic on biases and cognition in the intelligence profession, Heuer, Psychology of Intelligence.

39. To bring this point to full light, considering a possible God as omniscient, this entity should be also assumed as uncapable of any surprise. Only assuming that God could be surprised, considering all the direct access to knowledge about future events, it is possible to assume that, indeed, there is, after all, an absolute kind of surprise. However, here there is no need to assume such scenario as interesting for our purposes, although it could be in different contexts.

40. For the afficionados of philosophical formulations: (Sr) A subject S is surprised by an event e at time t₂ if and only if S has expected non-e at t₁.

41. What I really meant is that the child does not have category-expectations, but he/she has some general expectations about the future. If somebody through a coin, it is reasonable to assume that he/she will have the general expectation “I believe that something is going to happen”. However, he does not have a specific category-expectation about the coin and what side of it will turn up. In this sense, the surprise is induced by the absence of category-expectation toward something that is unfolding. That’s why being in a child’s position is so critically dangerous: it blocks any further calculation on what can happen next because the calculus is stopped at the very beginning. I thank Ralf Lillbacka for his comments in this regard.

42. Terry Quist accurately framed this point in a comment to an early version of this paper: “So when there is no theory or law regarding the occurrence or recurrence of an event, the event is a surprise or at least a novelty. An event cannot be expected until it is part of a pattern of experience”.

43. Del Boca, Le guerre coloniali del fascismo.

44. Fortin, “When Soviets Launched Sputnik, C.I.A. Was Not Surprised”.

45. Mises, Human Action, 110–117.

46. We will come back to this important point in the conclusions.

47. Kotkin. Stalin, vol. II: Waiting for Hitler.

48. Obviously, these measurements depend on several factors, and they should not be taken too seriously, but at least they can convey the idea.

49. The difference is between general reliability and local reliability of our cognitive processes. If I am making a mistake calculating an algebraic formula, this does not lead to general unreliability of my cognitive process, but only to a local unreliability (see Goldman, Epistemology and Cognitition).

50. Wilson, Five Myths About Nuclear Weapons, chap. 1.

51. For a brief analysis see Lillbacka. “Schelling Traps as Drivers of Intelligence Failure”, 15–16.

52. Feyerabend, Against method.

53. Kuhn, The structure of scientific revolutions.

54. After all, defense is in favor to the attacker in terms of information and general objective of war, as recognized at least since Clausewitz (Clausewitz, On War).

55. I thank Terry Quist for this point.

56. Ivi, cit., 17.

Additional information

Notes on contributors

Giangiuseppe Pili

Giangiuseppe Pili is an Assistant Professor at James Madison University, Intelligence Analysis Program. He is an Associate Fellow at the Royal United Services Institute for Defense and Security Studies. He is also an external research member of Intelligence Lab – Calabria University and a former lecturer in the International Master in Security, Intelligence and Strategic Studies at Dublin City University in Ireland. He earned a Ph.D. in philosophy from Università Vita-Salute San Raffaele (Milan, Italy). He is an editorial board member of the Italian Society of Intelligence and host of the series “Intelligence & Interview”. Along with Professor Mario Caligiuri, he is the author of the book Intelligence Studies (2020), which was included in the official digest of the Italian Secret Services (DIS).