GG Only - Denial & Deception (1)

Chapter 5

THE RESEARCH STUDY

“I haue already found a stratageme, To sound the bottome of this doubtfull theame.”

–Thomas Kyd, The Spanish Tragedie (1592),

Act II, Sc. 1.

So far I have recounted the history of military deception and surprise. I have also put forth a rudimentary general theory of stratagem. To verify it in detail would require a rather large-scale research project. However, as the relationship between surprise and deception has not previously received any re- search attention, even an exploratory study has value and interest. This chapter describes my own preliminary research and presents my empirical findings and conclusions.

At first sight, it seems odd that the strong tradition of the quantitative comparative case method inaugurated by the great German 19th century mili- tary historian, Hans Delbrück (1848-1929),1 and carried forward by the British Quaker physicist Lewis Richardson (1881-1953)2 and the American lawyer Quincy Wright (b. 1890)3 has not yet addressed itself to the factors of surprise or de- ception. On reflection, however, the answer seems clear. Those scholars dealt mainly with directly measurable and numerable categories. They laboriously compiled and systematically compared such characteristics as size of armies, populations, budgets, and national incomes; numbers of casualties, allies, or neighboring countries; and frequency of conflicts. By concentrating on such quantitative categories they slighted the “psychological” factors in war.4

However, even inherently qualitative factors such as surprise and deception can easily be manipulated quantitatively by characterizing or categorizing them

1See Gordon A. Craig, “Delbrück: The Military Historian,” in Earle (43), 260-283. 2Lewis F. Richardson, Statistics of Deadly Quarrels (Pittsburgh: Boxwood Press, 1960). 3Quincy Wright, A Study of War (2nd ed., Chicago: University of Chicago Press, 1965).

The first edition appeared in 1942. 4Although Richardson and Wright and perhaps even Delbrück would be among the first

to acknowledge the importance, even preponderance, of the psychological dimension.

81

82 CHAPTER 5. THE RESEARCH STUDY

in terms of their presence or absence and then counting the frequency of their occurrence across a number of cases or against other categories, which them- selves may be qualitative or quantitative in nature. Thus we can investigate the relative frequency of surprise, the coincidence of surprise with deception, the association between crying “wolf” and the attenuation of alertness, etc., etc.

5.1 The Case Studies

The study covers an inventory of all 68 cases known to me of strategic surprise and/or strategic deception that initiated a war, campaign, or major battle dur- ing the period 1914-1968 in connection with international war. (Also included are 47 selected examples or tactical surprise and/or tactical deception as well as 53 examples of military operations explicitly known to have not involved either surprise or deception.) The six specific defining criteria are:

1. By “surprise” I mean those instances where a sudden military action by one antagonist has not been predicted, much less anticipated, by its in- tended victim.5 Two modes are recognized: “strategic” and “tactical.”

2. By “deception” I mean an act intended, by its perpetrator to dupe or mis- lead a victim. Note that this definition excludes those instances where the victim unilaterally misunderstands. Two modes are recognized: “strate- gic” and “tactical.”

3. “Strategic surprise” is distinguished from “tactical surprise” by the degree to which the military action affects the victim’s mobilization, deployments, or grand strategy. In general, “tactical surprise” grades into “strategic surprise” in that region of the tactical-strategic scale where the locus of command shifts from the narrow zone of battle with its field commanders to directly involve other regions and higher military or political leaders.6

4. By “cases known to me” I mean those instances of surprise and decep- tion, explicitly identified as such, found in surveying the literature cited in the footnotes and bibliography of this book. (My “inventory” of cases of strategic surprise and/or strategic deception is very probably almost complete as all “strategic” operations mentioned in the standard and of- ficial histories surveyed carried at least some evidence or indication as to whether or not surprise or deception were involved.)

5. By “initiated a war, campaign, or major battle” I mean a deception and/or surprise that occurred at the opening phase of a war, campaign, or large- scale battle. This criterion was introduced to exclude those important

5For my belated or post-research thoughts on a more refined typology see Chapter VI, Part A (“The Variety and Intensity of Surprise”).

6For my suggestions for a more refined categorization see Chapter VI (“The Strategic- Tactical Dimension”).

5.1. THE CASE STUDIES 83

World War I 1914-1918 Soviet-Polish War 1920 Greco-Turkish War 1921-1922 Italo-Ethiopian War 1935-1936 Spanish Civil War 1936-1939 Sino-Japanese War 1937-1939 Soviet-Japanese Border Conflicts 1938-1939 Russo-Finnish War 1939-1940 World War II 1939-1945 Israeli War of Independence 1947-1949 Korean War 1950-1953 Vietnam Revolution 1946-1954 Suez War 1956 Bay of Pigs 1961 The Six-Day (Arab-Israeli) War 1967 Czech Invasion 1968

Table 5.1: The Sixteen Wars That Comprise The Census, 1914-1968.

but difficult-to-identify instances of follow-on or ongoing and usually im- provised exploitation drives in which the “dummy” of stratagem is “sold” right and left along the axis of advance.

6. Sixteen wars comprise the census for the period 1914-1968. They are listed in Table 5.1.

Obviously this list does not exhaust the large-scale martial conflicts of this period. But it does–I believe–include all confrontations that contain examples of the type of strategic surprise or deception that meet the specified census criteria.

One point should be stressed about the above definitions. Deception is defined (and used throughout this paper) only in its active sense; that is, as viewed by the practitioner rather than in terms of its effect on the intended victim. Thus I exclude Goethe’s aphorism that:

We are never deceived. We deceive ourselves.

Conversely, surprise is defined exclusively as viewed by the victim, in other words, as an effect. Consequently, the two terms–deception and surprise–are made to fall into quite separate logical categories: the perpetrator’s intention to deceive and the victim’s perception of surprise. Thus defined, these terms thereby avoid the easy pitfall of circular reasoning.

Because the distinction between “strategic” and “tactical” is vague, several of the cases arbitrarily included in my inventory of strategic surprises and de- ceptions can be challenged. I suppose the more vulnerable ones are Tanga (A1),

84 CHAPTER 5. THE RESEARCH STUDY

Dakar (A24), Alamein (A35), Tarawa (A40), Anzio (A41), Luzon (A52), and Kojo (A61), all of which were more-or-less localized operations whose wider im- plications are open to debate. However I cannot follow Professor Samuel Eliot Morison, who would reduce even OVERLORD to a case of mere “tactical” sur- prise on the niggling ground that the enemy expected invasion, even if he did not know the time, exact beaches, or strength of the blow. Morison simply misses the subtlety of the OVERLORD deception plans that gained just the right type of strategic surprise.7

There are several borderline cases of major battles involving surprise or deception that have been excluded because they do seem clearly to fall on the “tactical” side of the “strategic-tactical” criterion. Those examples of which I am aware are summarized together with the other known “tactical” operations in List B, and all receive more or less detailed description in the case study appendix. However, as these examples form neither a statistical sample nor a complete inventory, they will be discussed only illustratively in the main text.

I have definitely excluded all cases at the extreme of the small-scale, local actions of the commando type, although many of these achieved surprise and were intended as diversionary raids or served political-strategic ends. Indeed, General Telford Taylor is prepared to class Skorzeny’s spectacular kidnapping of Mussolini in September 1943 as a “strategic coup,”8 apparently because of its political character and in spite of Il Duce’s dissipated political and even symbolic significance. A good example of a border-line case that I have counted as “tactical” is the Dieppe “reconnaissance in force” in 1942 (Example B26). Of all 168 cases and examples studied, only two involved an initial attacking force of less than division size (or 5,000 men).9

There is one interesting “case” of strategic surprise credited in the litera- ture10 that I have excluded because it originated in the manufactory of a no- torious forger.11 This is the story told in a sensational book by the mythical Soviet defector “Colonel Kyril Kalinov” –actually the figment of the imagination of Gregori Besedovsky or his onetime assistant, Victor Alexandrov. Kalinov- Besedovsky “quotes” Marshal Zhukov as stating (in 1949) that the Germans were tricked into resuming their costly winter offensive against Moscow on 15 November 1941. Indeed, Hitler and the Army General Staff had decided to launch this final drive because they had grossly underestimated the Russian strength. Zhukov-Kalinov states:12

In this connection I can now disclose an important detail which has hitherto been kept secret. The report about the allegedly destroyed

7A detailed discussion of this typological problem is supplied in Chapter VI (“The Variety and Intensity of Surprise”).

8Brigadier General Telford Taylor, “Foreword,” in Charles Foley, Commando Extraordi- nary (New York: Putnam’s, 1955).

9Case A65 and Example B25. 10The credulous historian is Carell (64), 163-165 11Cyrille D. Kalinov (pseud.), Les Marèchau soviètiques vous parlent. . . (Paris: Stock,

1950). 12Kalinov (50), as translated in Carell (64), 163-164.

5.1. THE CASE STUDIES 85

330 divisions was launched by us deliberately to find its way to Germany through the Military Attachè of a neutral country whom we knew to be in touch with Germany’s military intelligence service.

5.1.1 The “Soft Underbelly” Policy

Was Churchill’s famous and controversial “soft under-belly” policy a case of strategic deception? I am unable to judge, and the plethora of sources and studies generate more fog than clarification. For example, the most detailed monograph on the general question of Churchill’s “second front” entirely evades its Balkan aspect.13 Indeed, I know of only one study that even begins to pose the relevant questions.14

We know that throughout the war Churchill did advocate pressure on Ger- many through the Mediterranean, which he deemed to be its most vulnerable flank–its exposed “soft under-belly,” as he first termed it in 1942.15 However, his policy varied in content, priority, and strength of advocacy in response to changing political-military considerations, particularly the collapse of Italy and the intransigence of his Russian and American allies who were pressing for a cross-Channel attack at the earliest possible moment. Churchill, however, ar- gued that such a direct frontal assault on the point of the enemy’s greatest strength would yield prohibitive casualties and likely defeat unless either Ger- many showed signs of breaking or the Anglo-American invasion force could be of such size that it would not risk a useless catastrophe at the beachhead–a second but grander “Dieppe.”16 Churchill fought hard and successfully for deferrals of the cross-Channel invasion from 1942 until 1943 and then, again, to 1944. Meanwhile he pursued a grand strategy of ever tightening economic blockade, increasing volume of strategic bombing, and stronger military pressure on the periphery, particularly the Mediterranean. His angry, critical Allies did raise the question of a hoax but only in terms of their suspicion that he was deceiving them about his intention to ever cross the Channel.17

Thus, the esoteric question of whether the “soft under-belly” was a desirable or even effective stratagem has been obscured by the still heated debate of whether it was a desirable or potentially effective strategy.

As already described, the British deception plan for the 1943 invasion of Sicily included a feint against Greece.18 But aside from this one instance, there is no direct evidence that the British made any further use of this ruse. However, I am struck by that fact that the Germans continued to the end to divert

13Trumbull Higgins, Winston Churchill and the Second Front (New York: Oxford Univer- sity Press, 1957). Even Higgins’ Soft Underbelly (New York: Macmillan, 1968) misses the deception possibility.

14W.G.F. Jackson, The Battle for Italy (New York: Harper & Row, 1967). 15In his note dated 12 November 1942 surveying the prospects in the Mediterranean. 16Example B26. 17See, for example, Ivan Maisky, Memoirs of a Soviet Ambassador, The War: 1939-43

(New York: Scribner’s, 1968), pp. 245-365; and Sir Samuel Hoare, Complacent Dictator (New York: Knopf, 1947), pp. 106-108.

18See Case A38.

86 CHAPTER 5. THE RESEARCH STUDY

N. Africa Eastern Norway Balkans & Italy West Front & Denmark Home Total

Date Ger/All. Ger/All. Ger/All. Ger/All. Ger/All. Ger/All. Ger/All. 1939 Sep 18 42/70 65/30 /5? 117/105? 1940 May 10 0/2 136/156 4/ /0 /0? 156/158 1941 Jun 22 7/0 2/3? 38/0 123/151 13/0 22/80? 205/234? 1942 Sep 169/ 1943 May 7 7/0 2/0 /0 /0

Jul 10 13/0 7/8 /0 /0 Sep 3 /0 17/19 /0 /0 Oct-Nov /0 19/19 /0 /330 /0 260/ Dec /0 18/13 /0 /0

1944 Jan 22 25/0 23/18 53/0 179/ 16/0 Mar /0 24/21 /0 /0 Apr 14 26/0 23/ 51/0 206?/ 17/0 0/ 336/ May 11 /0 23/25 /0 /0 Jun 6 28/0 26/25 59/7 165/ 18/0 0/31 296/ Aug 15/0 26/20 /0

1945 Jan 1 7/2 27/19 76/71 133/461 17/0 0/0? 260/553 Apr /3 23/17 147/

Table 5.2: Geographical Distribution of German and Allied Divisions in European Theatre. References: Jackson (67), 26-27, 317-320, 326-343; Collier (67) 386-387, 454; Guderian (52), 150; Bradley (51), 494; Butcher (46), 520-521; Erickson (62), 767-768; Ziemke (68), 7-8, 19, 370, 416; Whaley, Barbarossa (69), Chapter III, Tables I and 2; Deane (47), 145; Correspondence, I (57), 318.

precious strength to defend against a never-to-materialize amphibious invasion of the Balkans.19 The question therefore remains: Did the Allies–or rather, the British–do anything to deliberately feed the fires of Hitler’s expectations? If the answer can be shown to be “yes,” then we are dealing with one of the more brilliantly successful stratagems. Even if the answer is “no,” the case still serves as an excellent model of how a stratagem could have been applied at the level of grand strategy, because it did cause a major dislocation in the Wehrmacht’s deployments, without detracting from the strength of the cross-Channel attack when it did finally materialize.

A measure of the serious maldistribution of Germany’s flagging strength is seen in Table 5.2.

5.2 The Coincidence of Surprise with Deception

I have identified 68 cases of strategic surprise and/or strategic deception during the 55 year period from 1914 through 1968. A rough measure of the rarity of these events compared with the very large number of other major military operations during the same period is that they averaged only 1.3 per year; or, if we count only the 30 wartime years, 2.2 per year. However, when we consider that most of these represented major turning points in the fortunes of war–indeed, account for a significant proportion of such crucial events–the importance of research on the subject becomes clear.

Of the 68 “strategic” type cases, 61 involved strategic surprise and 57 in-

19Jackson (67), I, 138-139, 317-322; Majdalany (57), 22; Churchill, IV (50), 823-824; Ziemke (68), 132; Guderian (52), 310. The credulous Germans included Warlimont (64), 337, 352, 383-386, 499; Guderian (52), 310; and General Siegfried Westphal, The German Arm in the West (London: Cassell, 1951), pp. 150, 200.

5.2. THE COINCIDENCE OF SURPRISE WITH DECEPTION 87

Strategic Level Tactical Level (Type A Cases) (Type B Examples)

Number Percent Number Percent

Both Surprise & Deception 49 73.1% 25 53.2% Surprise only 11 16.4% 19 40.4% Deception only 7 10.5% 3 6.4%

Totals: 67* 100.0% 47 100.0%

Table 5.3: Deception and Surprise at Strategic and Tactical Levels. *One case (Case A67) was added too late to incorporate in all the quantitative analyses.

volved strategic deception. (These yield rates of some two strategic surprises and two strategic deceptions per year of the 30 years of actual warfare during the 55-year period studied.) Table 5.3 shows the distribution of these cases among their logical conditions.

Immediately we see that surprise and deception are commonly associated, particularly at the level of strategic operations (although still more often than not at the higher tactical levels). Is this frequent relationship something other than coincidence? Recall that these two categories have been defined so that deception could be a “cause”, and surprise an “effect.” Moreover, Table 5.3 verifies that the categories of surprise and deception have been defined so as to be logically quite distinct and are not merely tautologically linked. Thus the table shows 11 cases where the victim was surprised without the need of deception and 7 cases where deception was used but failed to yield surprise.

What, then, is the relationship between deception and surprise? Examining the 115 detailed accounts of operations in Appendix A, we find that surprise is a consequence of several factors. These factors may act singly or in concert to produce surprise. In no particular order of importance, these are:

Secrecy Tight security or ineffective enemy penetration does sometimes shield intentions or the clues pointing to intentions. The victim remains un- warned. However, this absolute condition does not exist in a single one of the cases. Some more-or-less specific warnings are present in all cases. Security and counterintelligence serve, at most, to protect only part of the truth. But this is sometimes enough to open the door to uncertainty. E.g., Cases A28 (Russia), A30 (Pearl Harbor), A57 (Korea), etc.

Preconception The victim of surprise forms an estimate of his opponent’s intentions and capabilities. This estimate is often wide of the mark. E.g., Cases A21 (France), A25 (Sidi Barani), A28 (Russia), A30 (Pearl Harbor), A45 (Normandy), A58 (Russian intervention in Korea), etc.

Deception The deliberate misleading of the victim.

Response Time There seem to be rare cases where the sheer physical ability of a victim to deploy to meet a fully perceived threat is so slow (relative to the attacker’s ability to quickly plan and deploy for attack) that the

88 CHAPTER 5. THE RESEARCH STUDY

Strategic Level (A) Tactical Level (B) War Dates Sur. Dec. Both Total Sur. Dec. Both Total WW I 1914-1918 1 0 9 10 10 0 9 19 Soviet-Polish 1920 1 0 0 1 - - - Greco-Turkish 1921-1922 0 0 1 1 1 0 0 1 Italo-Ethiopian 1935-1936 - - - 0 - - - 0? Spanish Civil War 1936-1939 3 0 0 3 2 0 0 2 Sino-Japanese 1937-1941 - - - 0 - - - 0? Russo-Jap. Border 1933-1939 0 0 1 1 0 0 0 0 WW II 1939-1945 5 5 30 40 5 3 15 23 Russo-Finnish 1939-1940 - - - 0 - - - 0? Israeli Ind. 1947-1949 0 0 1 1 1 0 1 2 Korean War 1950-1953 1 0 4 5 - - - 0? Vietnam Revltn. 1946-1954 0 1 0 1 - - - 0? Suez 1956 0 1 1 2 - - - 0? Bay of Pigs 1961 0 0 1 1 - - - 0 Six-Day War 1967 0 0 1 1 - - - 0? Totals: 11 7 49 67 19 3 25 57

Table 5.4: Incidence of Surprise and Deception in War, 1914 through 1963.

way to strategic victory is still open. Perhaps we should then speak of strategic paralysis or impotence rather than strategic surprise. E.g., Case A61 (the Kojo feint) and, just possibly, Case A59 (the Inch’on landing).

These four factors leading to strategic surprise–secrecy, preconception, de- ception, and response time–are, of course, quite gross categories. They are ex- amined more closely below. Moreover, they certainly do not exhaust the array of possible factors. For example, another that can well be imagined is salience: a powerful “victim” being simply uninterested in what some much weaker enemy might have in train, confident that any loss could always be regained. However, the four factors named are the only ones identified in the cases studied.

The incidence of surprise and deception among the 16 wars examined be- tween 1914 and 1968 (see Table 5.4) shows that they are fairly ubiquitous phenomena. Seemingly they are absent in only three of the wars, the Italo- Ethiopian, Sino-Japanese, and Russo-Finnish. (While I have confirmed their absence in the “strategic” operations of those wars, I believe more detailed re- search would turn up at least some instances of surprise at the “tactical” level.)

Similar conclusions are apparent in Table 5.5, which shows the frequencies of surprise and deception through time, decade-by-decade. This again shows, that these phenomena are very much a part of modern warfare.

Considering the last two tables together we get some hints about the nature of yet a third table, one which can, however, be generated only by knowing the absolute frequencies of battles in each war (and each decade). Only by doing this could we generate the table (or, rather, tables) showing the relative proportions (across wars and through time) between surprise and/or deception on the one hand and the absence of the characteristics on the other. However, given a rough, impressionistic grasp of the magnitudes involved, I would expect such a table to show the following important trends:

First, the proportion of military operations involving surprise unaided by deception is sharply declining. I would attribute this almost entirely to the

5.3. THE CAUSES OF SURPRISE 89

Period Surprise Deception Both Total 1914-1919 1 0 9 10 1920-1929 1 0 1 2 1930-1939 3 0 2 5 1940-1949 5 5 30 40 1950-1959 1 2 5 8 1960-1967 0 0 3 3 Totals: 11 7 50 68

Table 5.5: Strategic Operations (Type A Cases).

trend throughout our century in which intelligence analysis is improving at a much faster rate than are the techniques of passive security. Thus, modern surveillance and communications-ferreting systems and computer-aided analyt- ical techniques are rapidly increasing in their ability to disclose the physical evidence of military operations, while security countermeasures have not kept pace.

Second, the proportion of military operations involving deception alone–i.e., where stratagem has failed to yield surprise–has remained at a rather constant but low level. This I would attribute entirely to the rather generally poor level of understanding of the nature of stratagem. I presume that improved deception techniques would correct this deficiency.

Third the proportion of military operations involving both surprise and de- ception has sharply increased. The theory of stratagem would attribute this entirely to the trend in military doctrine toward various strategies of the “indi- rect approach.”

5.3 The Causes of Surprise

Writers on surprise in war will sometimes single out specific factors to which they attribute its achievement. At this point I will simply collate these factors and examine them in the light of the empirical case study evidence.

5.3.1 Terrain

There are certain terrain features that can conceal the movement of military forces. Frederick the Great was perhaps the first writer to explicitly emphasize this factor and to recommend the two obvious intelligence counter-measures: behind-enemy-lines espionage to extend vision beyond line-of-sight and long- range reconnaissance-in-force to acquire early contact and maintain it. This element of the “principle” of “security” is second-nature to professional soldiers. It will not be investigated in this paper, although detailed case or simulation studies might serve to identify its parameters.

90 CHAPTER 5. THE RESEARCH STUDY

5.3.2 Weather

Meteorological conditions are occasionally mentioned as important in gaining surprise, although it is striking how few are the official military historians who include this on their check-list of points to be covered.20 Indeed, the weather factor is one of those that is so very “obvious” that it is usually overlooked by war historians and analysts. The field commander must of course, be cognizant of weather, but some appalling examples are recorded where rear echelon planners and theater commanders remained unconcerned by even disastrous changes in weather.21

Two specific types of weather are relevant. First, there are those that render movement “impossible.” To the extent that these “impossible” conditions can be surmounted, surprise will follow. These are intense cold and mud-producing rainstorms.22

The other type of weather is that which conceals movement. Here night (darkness) and fog (or mist, rain, overcast, etc.) are the two relevant subtypes. If the attacker is trained to maneuver in such conditions, it is asserted that he can greatly increase his chances of surprise.23 In earlier times night attack was limited to small, highly trained, tactical teams, because of the virtual impos- sibility of controlling large groups of unseen (and unseeing) moving men. It was the speciality of the guerrilla and commando, although the masking dark has been mastered by some large units to conceal large scale redeployments.24

However, the introduction late in World War I of navigational and communi- cations gear unaffected by visibility made night attack a feasible technique to gain surprise at the level of “grand tactics” and strategy.25

Dividing the 24-hour day into three equal periods, we see that night at- tacks are, as predicted, more commonly associated with surprise than with non-surprise situations. Over half the cases of surprise began in the night, while this was true of only one-quarter the non-surprise cases as in Table 5.6.

Similar remarks apply to fog. In the late 1920’s, Captain Liddell Hart noted as a consequence of his encyclopedic study of World War I battles that the

20Thus, while the indices of the official British history of the Great War contain at least 34 separate references to the effects of fog and mist, they are thrown in rather haphazardly. See Hankey, I (61), 806-807.

21Thus, at the Battle of Loos on 25 September 1915, the Commander of the attacking British 2nd Division, Major-General Sir Henry Horne, insisted that Haig’s “programme must be carried out whatever the conditions.” The programme was that his division was to release chlorine gas prior to its assault. The conditions prompting his remark were that the wind had just switched toward the British trenches. The result was that his 2nd Division attacked with invalids who were promptly slaughtered by the ungassed Germans. Horne was promoted to command XVth Corps. See Liddell Hart (30), 192.

22For example, Ludendorff surprised the Czarist army by attacking in fearful cold on 7 February 1915 at the Masurian Lakes (Example B3).

23Liddell Hart was one of the first to develop a formal theoretical doctrine in his partially successful efforts to introduce “night attack” as a standard procedure in the British Army. See Liddell Hart, I (65), 212-218. Also Lt. Col. H.S. Yadev, Infantry Journal (India), No. 2, 1956, as digested inMilitary Review Vol. 38, No. 1 (April 1958), pp. 75-82.

24For example, the Chinese Communists, Viet Minh, and Viet Cong. 25Montgomery was a master of this and demanded that his troops trained accordingly.

5.3. THE CAUSES OF SURPRISE 91

Surprise No Surprise Time of Day Number Percent Number Percent

Night/Dawn (2200-0559) 49 52.7% 9 23.7% Morning (0600-1359) 37 39.8% 22 57.9% Afternoon/Eve (1400-2159) 7 7.5% 7 18.4%

Totals: 93 100.0% 38 100.0% No data: 7 22

Table 5.6: Effects of Time of Day on Surprise.

Surprise No Surprise Natural Climate Number Percent Number Percent

Fog, mist 15 21.1% 4 12.5% Rain, snow, cloudy 7 9.9% 7 21.9% Clear 49 69.0% 21 65.6%

Totals: 71 100.0% 32 100.0% No climate data 29 29 Not applicable 6 0

Table 5.7: Relationship Between Natural Climate and Surprise.

presence of fog was very often the only characteristic that served to distinguish surprise attacks from those met by alerted defenders.26 The relationship be- tween fog and surprise given by my own data is shown in Table 5.7.

This table shows that surprise is more commonly found associated with fog than those circumstances where surprise did not occur. However the data is both skimpy and already somewhat biased by the tendency in some of my sources to disproportionately report those instances where fog did contribute to surprise. Thus, while we should probably agree with Liddell Hart that fog often contributes to surprise, further research would have to be done to assess the magnitude of this effect. (Note also that, contrary to my expectations, rainy– and snowy or cloudy–conditions are quite negatively associated with surprise, at least on the basis of my weak data.)

Liddell Hart implies that his remarks about fog apply to localized (that is, tactical) battles only. Indeed, if we compare only those situations of strategic surprise with those of tactical surprise we find that our weak data again supports this. (Moreover, as we might suppose, inclement weather of all types favors tactical as opposed to strategic surprise.)

Included as a special meteorological factor is man-made weather as an aid to concealment of the attacker or befuddlement of the victim and consequent

26Liddell Hart, I (65), 216. He implies his conclusion was based on some sort of statistical analysis. If so he has not published it.

92 CHAPTER 5. THE RESEARCH STUDY

Strategic Surprise Tactical Surprise Natural Climate Number Percent Number Percent

Fog, mist 6 15.8% 9 27.3% Rain, snow, cloudy 3 7.9% 4 12.1% Clear 29 76.3% 20 60.6%

Totals: 38 100.0% 33 100.0%

Table 5.8: Cases of Natural Climate as Distributed Between Strategic and Tactical Surprise.

attainment of surprise. The innovation of smoke,27 gas,28 and artificial illumi- nation29 gave commanders enough environmental control to manipulate some tactical situations in World Wars I and II. The post-war period has seen im- provements in all these fields plus an important innovation in the last in the form of infra-red “sniperscopes.” And today we appear close to gaining a de- gree of climatic control (as opposed to prediction) sufficient to affect strategic surprise as well.30

My own data on artificial climate is very spotty. Thus Table 5.9 should be taken lightly, particularly as I have had to make the very dubious assumption that those cases where I found no data did not involve gas or smoke.

For what little it is worth, my empirical data shows that, in general, artificial climate has had only a trivial relationship to surprise. To the extent that there is any positive relationship, it applies only–as commonly assumed–to the poisonous or irritant gases. However, if we look only at the use of gas and smoke in the contexts of strategic and tactical surprise, we do see that they do–as surely expected–figure more prominently in tactical situations [see Table 5.10].

5.3.3 Timing

Third, there is the temporal factor. This has at least four aspects: speed, schedule changes, periodic, and associational.

Sheer speed is often quite sufficient to gain tactical and sometimes even strategic surprise. This truth has long been almost universally recognized, al- though discounted by those many who prefer to risk their fortunes in set-piece battles. These latter commanders and theoreticians might contemplate the ac- cidental finding of this study that only one commander was found who proved

27On smoke see Liddell Hart, I (65), 216-217; and Hankey, I (61), 229, 230, 246, 252, 300, 409.

28On gas see Leo P. Brophy and George J. B. Fisher, The Chemical Warfare Service: Organizing for War (Washington, D.C.: OCMH, 1959); and Farago (42), 276.

29On artificial light from World War I to the Korean War see Fuller (48), 413-415; Maskelyne (49); Hart, I (65), 217-218, 226, 297, 380; and Futrell (61), 421.

30Herman Kahn and Anthony J. Wiener, “Technological Innovation and the Future of Strategic Warfare,” Astronautics and Aeronautics, Vol. 3, No. 12 (December 1967), pp. 28-48.

5.3. THE CAUSES OF SURPRISE 93

Surprise No Surprise Artificial Climate Number Percent Number Percent

Gas and smoke 1 1.0% 2 3.2% Gas only 8 8.1% 3 4.9% Smoke only 4 4.0% 5 8.2% Neither 26 26.3% 8 13.2% No data (presumed neither) 60 60.6% 43 70.5%

Totals: 99 100.0% 61 100.0% Not applicable: 7 0

Data in Simplified Form

Gas and/or smoke 13 13.1% 10 16.4% Neither (incl. no data) 86 86.9% 51 83.6%

Totals: 99 100.0% 61 100.0%

Table 5.9: Relationship Between Artificial Climate and Surprise. Data is shown in full (top) and simplified (bottom) formats.

Strategic Surprise Tactical Surprise Artificial Climate Number Percent Number Percent

Gas and smoke 0 0.0% 1 2.4% Gas only 2 3.2% 6 14.3% Smoke only 2 3.2% 2 4.8% Neither 9 14.5% 17 40.5% No data 49 79.1% 16 38.0%

Totals: 62 100.0% 42 100.0%

Table 5.10: Cases of Artificial Climate as Distributed Between Strategic and Tactical Surprise.

94 CHAPTER 5. THE RESEARCH STUDY

Type A Type B Type C Cases Examples Examples

Advanced once 4 3 1 On Schedule* 26 23 34 Deferred once 16 10 13 Deferred twice 10 6 2 Deferred thrice 4 1 0 Deferred 4 times 1 1 2 Deferred 9 times 0 1 0 Deferred 16-21 times 3 0 1 Not applicable 3 2 0

Totals: 67 47 53

Table 5.11: D-DAY Schedule Slippage. *Note: All “no data” and “uncertain” cases have been given the benefit of the doubt by being classified as “On Schedule.”

consistently victorious despite his preference for the set-piece.31 This was Mont- gomery, who not only always enjoyed overwhelming preponderance in men and materiel but often obtained surprise by means other than speed. It has taken the nuclear threat of sudden, definitive attack to raise the salience of response- time to a point where systematic research has been applied, at least to that one specific environment.

D-day deadlines are often missed (Table 5.11). Among all 162 cases and examples where deadlines applied, almost half (or 43.9%) were delayed, 8 (4.9%) were able to go in ahead of their original schedule, and only slightly more than half (51.2%) remained on schedule.

In other words, put most conservatively, at least one-third of all military operations of, say, corps size and above are, for various reasons, unable to meet their deadlines. Moreover, this proportion holds roughly true regardless of the period, army, or size of operation studied. The two main reasons are weather (for example, amphibious and air operations schedules are highly dependent on the vagaries of weather) and administrative inability.

Curiously, this fact–that there is a more-or-less consistent 35-50% chance that a given attack schedule will not be met–is overlooked by both the historians and practitioners of war. It is highly relevant to our topic of surprise because this means that commanders and their intelligence services generally overlook this very likely contingency when estimating their enemies’ probable course of action, specifically the timing of attacks. Indeed, this oversight undoubtedly accounts for the rather noteworthy coincidence between schedule changes and surprise seen in Table 5.12

Table 5.12 shows three interesting facts. First, as expected, punctuality is not advisable if one seeks surprise. While its incidence was only 47.5% in

31A similar finding emerged–also quite unexpectedly–in my study of Guerrilla Communica- tions (67). The question of the trade-offs between these two types of battle deserves systematic research.

5.3. THE CAUSES OF SURPRISE 95

Surprise No Surprise Schedule Number Percent Number Percent

On Schedule 48 47.5% 35 57.4% Schedule Deferred 46 45.6% 25 41.0% Schedule Advanced 7 6.9% 1 1.6%

Total: 101 100.0% 61 100.0%

Table 5.12: Relationships Between Schedule and Surprise.

Day A B C Total Saturday 6 6 8 20 Sunday 10 4 11 25 Monday 13 9 9 31 Tuesday 10 6 10 26 Wednesday 8 5 5 18 Thursday 6 12 9 27 Friday 9 4 1 14 Not Applicable 5 1 0 6 Totals: 67 47 53 161 Mean Average: 8.9 6.6 7.6 22.1

Table 5.13: Days on Which Operations Were Opened, As Distributed Among Three Types of Cases and Examples.

situations of surprise, it jumps 10% (to 57.4%) in situations where surprise did not obtain. Second, as might have been expected (although I did not do so), of the two types (directions) of change the one more favorably associated with surprise is the advanced schedule, putting the attack in ahead of the original schedule. I presume this results simply from the enemy having less time to learn anything about the threatening attack. (This speculation is capable of empirical testing, unfortunately I did not think to collect data at that level of specificity.) Third, there is the ironic finding that even the annoying delays of administrative inefficiency and acts of God are rewarded by surprise, albeit with only a marginally greater chance.

The timing of surprise has minor but at least historically significant periodic aspects. Most human activities are geared in one way or another to daily, monthly, and seasonal cycles. Grounded in man’s physiological links with such natural cycles, they represent the unbidden influence of his physiology on his psychology, customs, and society.

Let us begin by looking at the periodicity of the days of the week (Table 5.13).

No particularly sharp pattern is evident in Table 5.13, which distributes the days on which operations were opened among the three types of cases and exam- ples. However, the wide divergence of several “cells” from chance expectation

96 CHAPTER 5. THE RESEARCH STUDY

(their mean average) bears a closer look. It is widely believed that in World War II the Axis powers (or at least

Germany and Japan) went out of their way to attack on the Sabbath (specifically the Christian Sunday). This is taken as proof of their devious, Godless, and monstrous nature. Moreover, it implies the plausible hypothesis that Sabbath (or weekend) attacks will tend to catch the victim at rest or worship. This specific popular belief and the underlying general hypothesis are easily tested, requiring only collation of the readily available dates of attack and a check of calendars to determine the day of week. Had any military historian done this he would have discovered that both the belief and the hypothesis are false.

The Nazis did not show any special preference for either Sunday or weekend attack. Of 24 attacks launched by them in World War II, only one-sixth (4) fell on Sunday and only one-eighth (3) on Saturday. This is virtually the same proportion as shown by the Germans in World War I, the British and Americans in World War II, and the Israelis (with a single Friday-Saturday Sabbath attack on Egypt) in the post-war period. Nor do the French, Japanese, or Italians show any noteworthy clustering about as particular day. The extreme preference for Sunday is shown by the Communist states (taking Soviet Russia, Red China, and North Korea) where even then only one-fourth of their attacks (3 out of 11) were started on Sunday. In general, no special bias for Sunday (or Sabbath or weekend) attack exists for any one nation or during any one period of history since 1914. Incidentally, the Pearl Harbor case is the only one where it is known that Sunday was deliberately chosen. Even then, the Japanese selected that day not so much for its day-of-rest inattentiveness as for the fact that it was the one day they could count on the U.S. capital ships–their target–being in port.32

Indeed, as Table 5.14 shows, the only clearly biased day is Friday which, oddly enough, is rather sharply avoided. Of the 161 cases studied only half the chance expectation are Fridays (14 instead of 23). I have no explanation for this. Moreover, the mystery is deepened by the fact that avoidance of military initiative on Friday was entirely a phenomenon of the First World War.33 Of 51 World War I cases studied, only one began on Friday. Thus, while I have accounted for this remarkable bias against Friday, I am quite unable to explain it. One reader34 has plausibly suggested superstitious avoidance–Friday being the Westerner’s traditional “unlucky day” and World War I a time when such superstitions were commonly heeded by soldiers.35

32See Case A30. 33Table not reproduced. 34Anonymous. 35This hypothesis, that Friday was avoided due to superstition would be confirmed by

finding explicit admissions in the memoirs of staff officers. The data does not support my own cynical suspicion that the senior generals might avoid upsetting their elaborate–indeed, sybaritic–weekend social plans by initiating military operations just before them. If this hy- pothesis were true, then one would expect Thursdays to also have been avoided and Mondays, Tuesdays, and Wednesdays to be the most common days for launching attacks. Neither was the case in World War I. In fact, as Thursdays, by appearing with twice its chance expecta- tion, entirely take up the slack of Fridays, this supports the reader’s hunch that Friday was indeed deliberately avoided and Thursday substituted for it. Chance remains a still possible but very lame explanation, one readily verified by taking an inventory or large random sample

5.3. THE CAUSES OF SURPRISE 97

Surprise No Surprise Day Number Percent Number Percent

Saturday 12 12.0% 8 13.1% Sunday 12 12.0% 13 21.3% Monday 21 21.0% 10 16.4% Tuesday 13 13.0% 13 21.3% Wednesday 12 12.0% 6 9.8% Thursday 18 18.0% 9 14.8% Friday 12 12.0% 2 3.3%

Totals: 100 100.0% 61 100.0% Mean Avg.: 14.3 8.7

Table 5.14: Relation Between Surprise and Day of Attack

Looking at the relationship between surprise and the day of week (Table 5.14), I was startled to find that the two extremes involve Sunday and, again, the somewhat mysterious Friday. Sunday is, in fact, the one day least associated with surprise; and Friday the one most closely linked.36

Perhaps the fact that Sunday is not favorably associated with surprise is simply because it is the mythically “expected” day. But, how can we account for Friday as the preeminent day of surprise? We cannot dismiss it as part of the off-duty weekend because the Friday attacks came in the morning or during the work day before the traditional weekend relaxation. Nor can we relate it to the overall rarity of its occurrence, because elimination of the set of 51 World War I cases (with their one Friday, which incidentally was a surprise attack) makes Friday appear in the remaining cases with its normal chance frequency. Thus I have no explanation for the curious “predictions” that if one attacks on a Friday he quadruples his chances for surprise and if he attacks on a Sunday he halves his chances.

I am not aware of any special bias for either attacking or withholding attack on regular festive or high holy days such as Christmas or Easter. In either case, an answer would require that a much larger number of cases be available or, failing that, explicit information about such considerations in the planning process. As it is, the only specific case I have found is that one of the many deferrals of his invasion of France (the one from December 17th to January first) was made by Hitler ostensibly to permit his troops Christmas leave.37

A special consideration–particularly among Far Eastern armies–is a propen- sity to schedule the taking of objectives–as opposed to the launching of the attack–to celebrate the remembrance of some auspicious occasion. Thus, the Japanese in World War II had a regular custom of scheduling the capture of cities to coincide with their Emperor’s birthday. Such seeking of trophies to present as Imperial birthday gifts was, naturally enough, usually unsuccessful,

of World War I battles. 36As shown mare sharply by interchanging the axes of this table. 37Taylor (58), 60.

98 CHAPTER 5. THE RESEARCH STUDY

except in the case of easy targets such as Singapore.38 Even Douglas MacArthur played this game when through Xth Corps on 23 September 1950 he directed the 1st Marine Division–to the distress of its commander, Major General O. P. Smith who, correctly, calculated a high casualty rate for his men–to take Seoul on the 25th, which would mark 90 days from its capture. Somehow, MacArthur had been led to believe that 90 days held some magic significance for East Asians. In fact, Seoul was not secured until 1630 hours the 27th and 708 Marine battle casualties later; and MacArthur’s triumphal handing back of the city to a tearful Syngman Rhee was delayed to the 29th. However, as with MacArthur’s premature announcement of the recapture of Manila in 1944, the realities had not stayed his official declaration that the 25 September deadline had been met.39

5.3.4 Preconceptions

“American people . . . expect you to be able to say that a war will start next Tuesday at 5:32 p.m.”

–General Walter Bedell Smith

CIA Director-Designate, 195040

“Late in May [1941] one of my informants whose information had always proved true came to tell me that at 3 A.M. on the fourth Sunday in

June [i.e., 22 June] the Germans would march into Russia.” –Louis P. Lochner,

AP Bureau Chief, Berlin.41

Surprise also has certain associational or sequential preconditions and conse- quences. All I mean by this is that the expectations of a person or a group (such as an intelligence service or political-military decision makers) are in large part based on past experience. Because men cannot obliterate their memories (short of hypnosis, hysteria, brain damage, or death) their judgment of future events– including warnings–are very strongly biased against perceiving the developing patterns that signal the unprecedented. Consequently they are surprised when the event finally intrudes itself and their understanding of it is an ex post facto, hindsight, or retrospective process. Thus our understanding of such instances of strategic surprise as BARBAROSSA and Pearl Harbor come only from hind- sight recognition of the interplay of perceptions and stratagem. Conversely, the

38James Leasor, Singapore: The Battle That Changed the World (Garden City, N.Y.: Dou- bleday, 1968), pp. 259-260; and Farago (67), 206.

39Sheldon (68), 301, 311-312, 316; Heinl (68), 210-211, 239-240; Montross, II (55), 244, 264, 233; and Geer (52), 116. See also Appleman (61), 527, 532, 536.

40Time, 28 August 1950, p. 14. 41New York Times, 6 June 1942, page 4. Lochner’s informant is now known to have been

Hermann Maass who acquired it through the then retired Colonel-General Beck from a still- unidentified general in the OKR[?]. For my detailed reconstruction of this fascinating tale of liochyerrat [?]–one conveniently overlooked in most of the German Resistance literature–see Whaley, Operation BARBAROSSA (69), Chapt. III.

5.3. THE CAUSES OF SURPRISE 99

Surprise No Surprise False Alerts Number Percent Number Percent 0 85 83.3% 58 95.1% 1 8 7.9% 2 3.3% 2 or more 9 8.8 1 1.6% Total: 102 100.0% 61 100.0%

Table 5.15: Relationship Between False Alerts and Surprise

most assuredly successful stratagems are, I suppose, those that encourage a pat- tern of ambiguous signals, one of whose horns points in an expected but wrong direction.

It should be possible to adapt certain existing techniques to enable the in- telligence analyst to escape the bias of the preconceptions that obscure most stratagems.42 Such a serendipity machine (computer or manual) would, I pre- sume, involve a two-step process: a random (i.e., preconception-free) search for patterns in military posture followed by a comparison of such patterns with the array of theoretically possible military actions (i.e., capabilities). Those with best fit would be candidates for closer scrutiny and verification by conventional intelligence procedures.

The earlier, whimsical finding–that procrastination can help generate surprise– can be explained by the “cry-wolf” syndrome. There is some crude data to support this, as shown in Table 5.15.

Here we see that the instances of surprise are indeed somewhat more com- monly preceded by false alerts than in those instances not involving surprise. Moreover, the trend is that the greater the number of false alerts, the greater the chance of their being associated with surprise.

Examining the factor of preconception more closely, we can see that it is composed of many, complexly interrelated, elements: gullibility, prior history of the conflict (Case A51: Ardennes); “lessons” drawn from perceived analogies (Case A59: Inch’on Landing), salience or priority given the potential conflict in the face of competing problems (Case A18: Norway), the previous history of false alarms, etc. Let us take the last as an example (Table 5.16). This is the familiar “cry wolf” syndrome. It is present in 17 of the 102 instances of surprise. Moreover, its Aesopian moral seemingly holds in most or all of the 17 cases, the false alarms serving mainly to undermine the credibility of the source and dull the effect of subsequent warnings.

42Several papers on “probabilistic information processing systems” by Dr. Ward Edwards of the Institute of Science and Technology of the University of Michigan are directly related to this problem of how to overcome human preconception. I am indebted to Mr. William R. Harris for these references as well as for my awareness of the problem.

100 CHAPTER 5. THE RESEARCH STUDY

Surprise Achieved? Wolf Cried?: Yes No Yes 17 85 No 3 58

(N=163)

Table 5.16: Relationship Between Surprise and “Cry Wolf” Effect

Thus we see that the “cry wolf” syndrome constitutes an important pertur- bation in intelligence estimates of future enemy action. Moreover, it is ironic that in at least 5 of the 17 instances, some of the D-day warnings were quite au- thentic, the enemy having merely unexpectedly deferred the operation.43 The consequence was, of course, that several superb intelligence sources including Colonel Oster, Sorge, and Rössler received undeserved black marks on the eve of their subsequent definitive alerts.

43Cases A16, A19, A20, A21, A28.

5.4. SURPRISE AND CASUALTIES 101

5.4 Surprise and Casualties

“Battles are won by slaughter and manoeuvre. The greater the general, the more he contributes to manoeuvre, the less he

demands in slaughter.” –Churchill, The World Crisis, 1915 (1923), p. 5.

“Nearly all the battles which are regarded as masterpieces of the military art . . . have been battles of manoeuvre in which very often the enemy has found himself defeated by some novel expedient or

device, some queer, swift, unexpected thrust or stratagem. In many such battles the losses of the victors have been small.”

–Churchill, The World Crisis, 1915 (1923), p. 5.

“A big butcher’s bill was not necessarily evidence of good tactics.” –Wavell, telegram to Churchill, c. 20 August 1941.

Casualty statistics are one important measure of the costs of battle, and casualty ratios an index of relative success or failure of the antagonists. I have made extensive use of them in my analysis, because they are generally available. For example, I was able to find statistics on losses (in comparable categories) for both antagonists in 86 per cent of the combat actions studied.44

A favorable disproportion of casualty figures is frequently associated with surprise, so much so that it deserves a detailed statistical study to determine the relative roles of such other factors as stronger/weaker, attack/defense, etc. For the present, a cursory look will have to suffice.

For comparison let us glance at casualty figures in non-surprise situations.45

In the Great War the total battle casualties on the generally static Western Front were 4 million German to 7.4 million Allies, that is, “favoring” the Germans almost 1:2. The major offensives showed ratios generally between 1:1 and 1:2, usually favoring the defender. In World War II on the Eastern Front alone, Wehrmacht dead (all causes) totaled 3 to 3.5 million as against perhaps 12 to 13.6 million Russians, favoring the Germans about 1:4.46

While, it seems, no military writer has explicitly argued direct relationship between deception and low casualties, Shakespeare makes the assumption plain. Writing in 1599, he imagines King Henry the Fifth learning the casualties of the

44Specifically: 134 Both antagonists 15 One only 7 Neither

156 TOTAL COMBAT ACTIONS 7 Not applicable (no battle)

163 TOTAL MILITARY OPERATIONS 45For statistics on W.W. I casualties see Lieutenant-General Nicholas N. Golovine, The

Russian Army in the World War (New Haven: Yale University Press, 1931), pp. 75-104; and Willoughby (39), 131-135. For World War II see Esposito (64b), 399-401.

46Ziemke (68), 500.

102 CHAPTER 5. THE RESEARCH STUDY

Battle of Agincourt (1415).47 An English Herald reports the clear victory: 10,000 French slain to only 29 English.48 Henry piously exclaims:

. . . O God! thy arm was here; And not to us, but to thy arm alone, Ascribe we all. When, without stratagem, But in plain shock and even play of battle, Was ever known so great and little loss On one part and on the other? Take it, God, For it is none but thine!

In fact, the French casualties were almost certainly no less than 5,500 to something under 300 Britons, a quite plausible accounting by 5,000 well de- ployed English armor-piercing longbows against a chivalrous mob of 40,000 to 50,000. Moreover, the English were killing their prisoners that day.49 A copy- book case of the effects of technological surprise rather than divine intercession. While the real Agincourt is a relevant case, the point at issue is Shakespeare’s use of a legendary Agincourt. His 16th century was one of renaissance of the martial arts. National need, public concern, and the printing press combined to yield almost 200 military titles. The numerous English translations of classical texts included such advocates of surprise and stratagem as Frontinus (English translation 1539) and Vegetius (English translation 1572). And Machiavelli’s The Arte of Warre appeared in 1560. These translations and numerous origi- nal works–some discussing and recommending stratagem–were widely read by Queen Elizabeth’s subjects.50 And Shakespeare himself was immersed in this military ethos, if rather poorly read in its specialized literature.51 For example, 8 of his 15 uses of the word “stratagem” are explicitly in its military sense.52

A highly favorable casualty rate is one of the most valuable consequences of surprise. Tables 5.17 and 5.18 show this most emphatically.

Table 5.17 shows that the mean average casualty ratios favor the initiator of military operations by only 1-to-1.7 in non-surprise circumstances but by

47Shakespeare, Henry V, Act IV, Scene 8. 48Shakespeare merely copies these absurd figures from his main source for Henry V, the

second edition of the contemporary history by Raphael Holinshed, The Chronicles of England, Scotland, and Ireland (Second edition, enlarged: London: 1587), Vol. III, p . 533, as collated by W.G. Boswell-Stone, Shakespeare’s Holinshed (London: Lawrence and Bullen, 1896), pp. 195-196. See also Richard Hosley (editor), Shakespeare’s Holinshead (New York: Putnam, l968), p. 135.

49E.F. Jacob, The Fifteenth century, 1399-1485 (Oxford: Clarendon Press, 1961), pp. 153- 156. Note also that Holinshead himself also reports five to six hundred English killed as an alternative figure “of greater credit.”

50Henry J. Webb, Elizabethan Military Science (Madison: University of Wisconsin Press, 1965), pp. 3-16, 169-176; and C.G. Cruickshank, Elizabeth’s Army (2nd ed., Oxford: At the Clarendon Press, 1966), pp. 198-206.

51Paul A. Jorgensen, Shakespeare’s Military World (Berkeley: University of California Press, 1956); and Lily B. Campbell, Shakespeare’s “Histories” (San Marino, Calif.: The Huntington Library, 1947), pp. 255-305.

52John Bartlett, A New and Complete Concordance . . . of Shakespeare (London: Macmillan, 1894), p. 1476.

5.4. SURPRISE AND CASUALTIES 103

Number of Cases Average Casualty Ratio Surprise 87 1: 14.5 No Surprise 51 1: 1.7 Total 138 No data or N/A 29 Grand Total 167

Table 5.17: Effect of Surprise On Casualties, 1914-1967.

Number of Cases Average Casualty Ratio

Surprise 79 1: 5.3 No Surprise 45 1: 1.1

Total: 122 Deleted extreme *16 Grand Total: 138

Table 5.18: Effect of Surprise On Casualties In 90% Of Cases, 1914-1967. *Note: Sixteen rather than 14 were deleted only because this table has been summarized from a larger table where the 10% deletion rule was applied to each of the subcategories.

a thumping 1-to-14.5 when surprise is present. That is, surprise is more than eight times as effective at producing casualties.53 However, this almost fantastic difference is in large part the result of a small proportion of surprise operations that yielded extremely favorable casualty rates, largely as a result of mass sur- renders.54 While these extreme cases may very properly be credited as an effect of surprise, it will be more practical to limit our discussion to the more usual set of cases. This is accomplished in Table 5.18 merely by deleting the 10% most deviant cases, 5% at each extreme.

Table 5.18 shows that while the usual non-surprise operations produce ca- sualty ratios of about 1-to-1, those with surprise yield ratios of 5-to-1. That is, surprise may be rather reliably depended upon to quintuple the enemy’s casu- alty rates, relative to one’s own. This finding also holds roughly for both cases of “strategic” and “tactical” surprise.55

The theory of stratagem predicts that the more intense the surprise,56 the more favorable the casualty ratios. The empirical data quite emphatically veri- fies this, as seen in Table 5.19.

If, as the theory of stratagem asserts, deception is not only a main cause but also an enhancer of surprise, then the data must show that (a) casualty ratios are substantially greater in cases of surprise-with-deception than for those of surprise-sans-deception and (b) there is slight or no greater casualty ratios

53Incidentally, the ratios would have been still much greater had I totaled the actual casu- alties and then taken their mean averages rather than merely averaging the average ratios.

54For example, Case A27 and Example B30. 55Strategic surprise giving a 1-to-4.5 ratio and tactical a 1-to-6.2 ratio. Table not repro-

duced. 56For discussion of my index of intensity of surprise see Section G, below.

104 CHAPTER 5. THE RESEARCH STUDY

Intensity Average Casualty Ratios 0 1: 1.1 1 1: 1.7 2 1: 4.5 3 1: 5.4 4 1: 4.1 5 1: 11.5

Table 5.19: Effect Of Intensity Of Surprise On Casualties In 90% Of Cases, 1914-1967.

Number Average of Cases Casualty Ratios

Surprise with Deception 59 1: 6.3 Surprise without Deception 20 1: 2.0 No Surprise with Deception 5 1: 1.3 No Surprise without Deception 40 1: 1.1 Total: 122

Table 5.20: Relationships Between Surprise, Deception, and Casualties.

between cases of no surprise and those without deception. This is verified by Table 5.20.

Until now, I have described casualties only in terms of the ratio between those of the initiator of a military operation and those of his intended victim. But what of the absolute casualty figures? That is, what is the effect of surprise on actual casualties? While I have not generated the relevant tables, inspection of the raw data summarized in the Analytical Lists comprising Appendix B shows that the effect is two-fold. First, surprise increases the enemy’s total casualties, particularly in numbers of prisoners taken. The effect is greatly enhanced if the protagonist is prepared to exploit the initial opportunities opened up by surprise. Second, surprise decreases the protagonist’s total casualties. This effect is compounded of a direct consequence of surprise itself, which renders the enemy’s response confused and hence inefficient and of an indirect consequence of the speed associated with surprise, as in the Six-Day War (Case A66) where daily casualty rates were high but the whole operation ended quickly in a decisive victory at a far below expected cost. In sum, the overall military effect of surprise is to produce both relative and absolute figures that are favorable to the protagonist. Moreover, the long-range demographic or social costs are favorable to both sides. Again the relative costs are less for the protagonist, but here the absolute costs for both sides are less in killed, maimed, and military and civilian property damage.

5.5. SURPRISE IN OFFENSIVE-DEFENSIVE MODES 105

5.5 Surprise in Offensive-Defensive Modes

One of the longest standing debates in military theory has revolved about the relative merits of the offensive versus the defensive modes of battle. Changes in both weaponry and doctrine have shifted preference or emphasis first to one then to the other mode. While at the grossest level of analysis the debaters divide neatly between the advocates of attack and the advocates of defense, the issue becomes quite muddied at the next lower level of abstraction. There we find great variety: The defensive may be static (depending on forts or trenches) or mobile (depending on counterattack). The offensive may be conducted by crushing frontal attrition, or breakthrough followed by either the classical forms of exploitation or the Blitzkrieg’s “expanding torrent,” or various forms of out- flanking and other indirect maneuvers.

One of the legacies of the machine-gun and artillery-dominated Western Front was a “magic” number: three. Officers came to adopt the notion that a three-to-one local superiority in the zone of an offensive was necessary for a successful attack. And, indeed, this rule of thumb was consistent with statistical studies of the overall averages of Western Front battles.57 This bit of Great War doctrine was carried into World War II (and beyond). Its main exponents have included General Alexander and his two Eighth Army Commanders (Generals Montgomery and Leese) and his Fifth Army Commander (General Clark).58

Analyses seemed to show that even greater than 3-to-l superiority was needed to success of break-out attacks in such confined pockets as Normandy.59

Most Soviet and many American commanders also backed this doctrine. Even Chinese Communist military commanders–at least during the Korean War–are alleged to be “wary of launching an offensive unless their forces out- number the enemy by three to one” and only then under cover of night, if possible.60

Such magic numbers do have their origin in some limited or local experience. However, they become too readily generalized and applied without verification to only vaguely analogous situations. The ubiquitous ten-soldiers-to-defeat-one guerrilla is the most familiar recent example, and it is immediately exploded by comparative case studies.61

How does this 3-to-1 doctrine look in the light of our data on surprise? First, let us look only at the World War I data. Of the 47 offensive battles

for which data was available, the average strength ratio of attacker to defender was 2.1 to 1 (Table 5.21).

57See, for example, Lieut.-Colonel F.O. Miksche, Atomic Weapons and Armies (New York: Praeger, 1955), p. 114.

58See, for example, Majdalany (57), 254. 59Liddell Hart, “Lessons of Normandy,” unpublished(?) paper quoted in Miksche (55), 106. 60Andrew Geer, The New Breed: The Story of the U.S. Marines in Korea (New York:

Harper, 1952), p. 221. 61See, for example, Andrew R. Molnar, Undergrounds in Insurgent, Revolutionary, and Re-

sistance Warfare (Washington, D.C.: The. American University, Special Operations Research Office, 1963).

106 CHAPTER 5. THE RESEARCH STUDY

Surprise No Surprise Achievement Number Force Ratio Number Force Ratio

Victory (V+) 7 2.1 :1 0 – About as planned (V) 10 1.6 :1 5 2.1 :1 Below expectations (V-) 5 2.8 :1 5 1.3 :1 Defeat (D) 3 5.9 :1 12 1.8 :1

Table 5.21: Force Used to Gain Objectives In World War I. This table excludes the one extreme instance (Case A1) where the British enjoyed a 114-to-1 advantage and surprise at Tanga but failed to exploit either. Two other examples (B7 and B8) were excluded because they ware purely defensive.

Surprise No Surprise Achievement Number Force Ratio Number Force Ratio

Victory 18 1.2 :1 1 2.5 :1 About as Planned 28 1.1 :1 4 1.4 :1 Below Expectations 17 1.4 :1 9 1.4 :1 Defeat 4 1.0 :1 20 0.9 :1

Table 5.22: Force Used to Gain Objectives After World War I. Case A65 (Bay of Pigs) was deleted as an extreme figure, and for Example B29 (Mèdinine) was transposed from British defensive to German offensive.

There is no obvious pattern here. But, before comment, observe the compa- rable table for the data after the Great War (Table 5.22).

Both tables 5.21 and 5.22, if taken together, support two important conclu- sions. First, we see strong proof for the notion that a substantial superiority of force is needed, although my data yields force ratios of about 2-to-1 rather than the traditional 3-to-1. However, this general remark holds true only for the more usual type of offensive military operation, that is, the one not involving surprise. For that type there is a fairly direct relationship between force and degree of success, the more the force the greater the success. Second, and both more important and rather unexpected, is that surprise intervenes to shatter the direct and simple relationship between force and success.

5.6 Military Environments: Land, Sea, Air, Amphibious

Each of the three gross geographical environments–land, sea, and air–poses its own problems and opportunities for surprise attack. These are mirrored in the separate doctrinal traditions evolved by each of the three services–army, naval and airforces–of each nation and their various combinations by the hybrid doctrine of combined or amphibious operations.

Although I had not set out to limit my study of surprise and deception to land operations, the great majority of cases in the period 1914-1963 have, in fact, been of that type, as there have been relatively few large-scale battles since

5.6. ENVIRONMENTS: LAND, SEA, AIR, AMPHIBIOUS 107

Combat Environment Number Land 125 Amphibious 33 Air 4 Naval-Air 3 All naval 1 Airborne 1 Total: 167

Table 5.23: Distribution of Combat Cases Among Environments

Land Amphibious Other No. Percent No. Percent No. Percent

Surprise with Deception 46 36.8 25* 75.7 5 55.6

Surprise only 26 20.8 3 9.1 1 11.1 Deception only 4 3.2 3 9.1 1 11.1 Neither 49 39.2 2 6.1 2 22.2

Totals: 125 100.0 33 100.0 9 100.0

Table 5.24: Relationship of Combat Environment with Surprise and Deception, 1914- 1967. *Note: Includes two cases (A22 and A23) where deception succeeded but where surprise was not applicable.

1914 that were predominantly naval or air engagements. Fortunately, therefore, the number of cases involving combined operations has been large enough to cast light not only on that particular type but on its component environmental elements as well (see Table 5.23).

I will begin with a broad look at the relationship of military environment to surprise and deception (see Table 5.24). To do so I have included all cases. How- ever, the specific percentage figures that result should be taken lightly, because the category where neither surprise nor deception was present is grossly under- represented. In other words, the trends across columns are more meaningful than the percentages within columns.

Table 5.24 shows that, in the 20th century at least, surprise is present in almost all (85%) of amphibious landings, but in only about half (58%) of the land and other environments. The reason would appear to be not so much that the littoral environment offers greater opportunity for “natural or spontaneous” surprise as it is that deception is simply used in a larger proportion of such enterprises. Thus, the same table also shows that while deception was present in only 40% of the land operations studied, it was present in 85% of the amphibious cases.

The military environments can appear–and have done so from time to time– in their three unalloyed forms and in four combinations, seven modes in all [see Table 5.25].

108 CHAPTER 5. THE RESEARCH STUDY

Environment Military Mode Introduced Land Land War, Ground Operations Prehistory Sea Naval Warfare Antiquity Land-Sea Amphibious (or Combined, Antiquity

or Conjunct) Operations Air Aerial Warfare(including W.W. I (1915)

air-ground support) Land-Air Vertical Envelopment or Airborne Spanish Civil

Operations (glider, War (1936) parachute, helicopter)

Air-Sea Carrier Warfare W.W. II (1940) Land-Sea-Air Combined (or Amphibious) Operations W.W. II (1940)

Table 5.25: Military Environments and Military Modes.

Land Operations World War I World War II

Number Percent Number Percent

Surprise with Deception 12 26.1% 27 54.0% Surprise only 10 21.7% 7 14.0% Deception only 0 0.0% 3 6.0% Neither 24 52.2% 13 26.0%

Totals: 46 100.0% 50 100.0%

Table 5.26: Stratagem Trends in Land Warfare

As each of these types of operation offers unique opportunities for surprise and deception, the fact that they appeared at different points in history and tend to occur in different proportions of frequency in any given military epoch implies that some epochs should offer greater opportunity for surprise than others. While systematic empirical verification of this hypothesis is beyond the scope of the present paper, it is worth raising because it suggests a major conclusion. Namely, that in the race between the increased number of opportunities for surprise (i.e., the greater number of feasible optional means and paths to a given goal) and the increased effectiveness (and number of means) of intelligence warning systems in preventing surprise, the rôle of deception becomes ever more decisive in providing the breakthrough to surprise. This is borne out by the trend in land warfare from World War I to World War II, as seen in Table 5.26.

Here we see a sharp increase with time in the proportion of all cases of surprise, from 48% in the Great War to 68% in World War II. From this it could seem that surprise has become “easier” to achieve. In fact, the table also shows that this could be entirely accounted for by the parallel rise in the use of deception.

5.6. ENVIRONMENTS: LAND, SEA, AIR, AMPHIBIOUS 109

5.6.1 Land

Land battles and campaigns have generally covered the great majority of large- scale military actions throughout history. This is true even today when air and amphibious operations have come to play a highly important role. This circumstance is fairly represented in my study where, of 168 operations, 126 were entirely on land. As the bulk of this paper has been concerned with land warfare, I will not single that topic out for any special remarks here.

5.6.2 Sea

My superficial examination of the writings of such naval theoreticians as Admiral Mahan indicates that that rather considerable body of literature has progressed even less than that of land warfare in imbedding the concepts of surprise and deception into general theory. The immediate reason for this, as British Vice- Admiral Gretton has shown in a damning appraisal, is that naval theoreticians from Malian to Bernard Brodie and Roskill have either uncritically borrowed or, at best, grudgingly accepted the “principles of war” as propounded by the land warriors.62 Moreover, they built upon the more sterile traditions of land war theory when surprise and deception were downgraded. This is unfortunate, because naval operations intrinsically lend themselves to abstract analysis more readily than land operations. This circumstance is implied by the fact that, compared to armies, navies have fewer major units (ships and bases) to deploy, a simpler and more centralized communication net, a more uniform media (the sea) on and in which to operate, and a smaller and more precisely defined set of potential objectives (other ships or coastal targets). Moreover, this circum- stance is demonstrated by the fact that the entire modern analytical technique of operations research (OR) originated as a solution to the destroyer-versus- submarine dilemma in World War I.

5.6.3 Air

Aerial warfare (even including the rare airborne operation), like naval operations and unlike land war, is well-suited to the sorts of abstract analysis in which concepts like surprise and stratagem easily fit. However, I find no evidence that this has produced any greater theoretical attention to these concepts much less any new thinking than already exists in the theory and doctrine of land war. Indeed while the principle of surprise is recognized in the writings of airmen, the technique of stratagem does not have any special place in air doctrine, at least not in the public literature.

There is good reason to presume that surprise and stratagem do figure promi- nently in the classified air doctrines of all major powers. This would arise from the fact–implicit if not always explicit in the air literature–that these are very immediate and practical issues. From World War I to the present, the evolu-

62Vice-Admiral Sir Peter Gretton, Maritime Strategy (New York: Praeger, 1965), pp. 21-24.

110 CHAPTER 5. THE RESEARCH STUDY

tion of air and aerospace strategy and doctrine has been intimately linked with problems of speed, interception, evasive maneuverability, and early warning.

The new combination of nuclear explosives with missile delivery systems has, by its potential for sudden decisive wars, raised the salience of surprise to an issue of survival itself, and it is recognized as such in the literature.63 However, I am not aware of any advances in the theory of strategic deception, as opposed to such tactical ruses as multiple warheads (MERVS), dummy warheads, and electronic countermeasures (ECM) to confuse the victim’s local defenses after an attack is launched.

5.6.4 Amphibious Operations

“Littoral War, when wisely prepared and discreetly conducted, is a terrible Sort of War.

Happy for that People who are Sovereigns enough of the Sea to put it in Execution!

For it comes like Thunder and lightening to some unprepared Part of the World.”

–Thomas More Molyneux, Conjunct Expeditions (1759).64

The combined (or amphibious or conjunct) operation is probably the most difficult single type of military operation, seeking as it does by very definition to meld the recalcitrant structures, maneuvers, and evolutions of armies, navies, and–lately–airforces. However, it promises the greatest chance of gaining sur- prise precisely because it combines two or more environments, each of which multiplies the number of options for surprise.

Combined operations are not only the most difficult to plan and mount, but they are also generally believed to be the most risky. Until the beachhead is secured and operations expand outward to land battle, the viability of the entire enterprise is doubtful; and the cost for failure is deemed to be nothing less than disaster. The clear appreciation of this high risk led, I think, to the early application of stratagem to that type of operation. There is nothing like the exigent threat of disaster to open the minds of planners to unorthodox solutions.

The British themselves acknowledge it was their desperate plight in 1940 that brought quick acceptance of such off-beat expedients as combined, “special,” and deception operations. In any case, stratagem became more-or-less standard procedure in amphibious operations in 1915 and has remained so ever since, as Table 5.27 shows.

This circumstance explains, I would suggest, why amphibious operations do not, in fact, fail as often or as badly as their detractors predict. Indeed, they

63See particularly Y. Harkabi, Nuclear War and Nuclear Peace (Jerusalem: Israel Program for Scientific Translations, 1966), pp. 41-51. This is a useful summary of the literature on nuclear surprise. Its interest is enhanced because the author was, as Chief of Israeli Military Intelligence during the Sinai Campaign of 1956, deeply involved in the practice of strategic and tactical deception.

64Quoted in Heinl (68), xix.

5.7. VARIETIES AND INTENSITIES OF DECEPTION 111

Amphibious Operations World War I World War II After No. % No. % No. %

Deception and surprise 5 83.3 17 73.0 3 75.0 Deception only 0 0.0 2 9.0 1 25.0 No deception, but surprise 1 16.7 2 9.0 0 0.0 No deception, no surprise 0 0.0 2 9.0 0 0.0

Totals: 6 100.0 23 100.0 4 100.0

Table 5.27: Stratagem in Amphibious Operations

Amphibious Operations With Deception No Deception

Number Percent Number Percent

Victory 6 21.4 1 20.0 About as planned 11 39.3 0 0.0 Below expectations 5 17.9 3 60.0 Defeat 6 21.4 1 20.0

Totals: 28 100.0 5 100.0

Table 5.28: Outcomes of Amphibious Operations

are generally less risky than land operations (Table 5.28).

5.7 The Varieties and Intensities of Surprise and Deception

Surprise is most commonly viewed in the military literature as an either-or quality: the victim is either surprised or he is not surprised. Up to this point, I have also found it convenient to treat it at this crudely abstract level. However, surprise may be viewed in greater detail as having at least two psychological dimensions: variety (or extent) and intensity (or depth). And both dimensions are manipulable by the alternative choice theory of stratagem. One dimension of surprise is its extent or variety–the different forms or modes that surprise may take. The first problem is how best to categorize these various forms. There is no shortage of suitable existing typologies. It is tempting to adapt one of chose already in use in the communications research field. For example, even a primitive model such as the “who-what-where-when-why” mnemonic of professional journalists would give a useful starting point. Similarly Professor Lasswell’s familiar: “Who says what in which channel to whom with what effect.”65 However, I have preferred to build a new set, using the bits and scraps already given in the military literature. These have the merit of familiarity and,

65Harold D. Lasswell, “The Structure and Function of Communication in Society,” in Lyman Bryson (editor), The Communication of Ideas (New York: Harper, 1948), p. 37.

112 CHAPTER 5. THE RESEARCH STUDY

with slight addition and modification, seen suitable for my purpose. Accordingly I will categorize the varieties of surprise as a set with five elements or modes: intention, time, place, strength, and style.

Intention in the sense I mean it here is the go/no go element. Is the war (or campaign or battle) even possible? Before the event such questions as the following were asked in all seriousness. Dare Hitler invade Poland or England or Russia? Dare Tojo make war on the United States? Does Hitler intend to preempt in Norway; or the Israelis in the Sinai? Are Britain and France merely preparing a bluff against Suez? Does Rommel even consider counterattacking at Mersa el Brega? Such questions relate to the basic intentions and decisions to wage war or peace, to bluff or act, to attack or defend, to escalate or surrender. They are the fundamental preferences and choices that determine whether a given war, campaign, or battle changes from possibility to reality. In one sense, intention is a precondition of the other varieties of surprise.

Time. Unexpectedness of time is a key form that surprise takes. It is usu- ally measured in minutes, hours, and days; also sometimes in weeks or months, although these longer time spans begin to blend into the more indefinite, even statically defined periods or eras that are covered by the above category of intention.

Place refers to the point or area threatened, or to the direction or axis of operations.

Strength refers to the amount of military force committed to the operation. It is usually measured, as I have done it in this study, by numbers of troops (or such aggregates as divisions or brigades), tanks, ships, planes, or missiles. Others measure it by the potential firepower (weight of naval broadside, tonnage of bombs, megatonnage of nuclear strike, etc.). Still others also attempt to add some index to take into account such qualitative factors as morale or efficiency.

Style. By style I refer to the form that the military operation takes, the fash- ion in which it is carried out. Thus one may be surprised by such stylis- tic things as the details of the operational plan as it unfolds, the types of weapons or weapons systems used (cavalry, tanks, chemical-biological, dive-bombers, machine guns, missiles, etc.), or the specific tactics (ex- panding torrent, frontal assault, static defense, etc.). This category is unusual in that it is often overlooked in the literature on surprise and in that it incorporates technological surprise (new weapons or weapons systems), which is almost always treated as a special category.

The factor of technological surprise has already been discussed in the text, and several examples are given in the appendices. Moreover, this particular mode of surprise is already well studied, as illustrated in the examples of poison gas at Second Ypres (Example B5), the tank at Cambrai (Example B13), and

5.7. VARIETIES AND INTENSITIES OF DECEPTION 113

the atomic bomb at Hiroshima (Case A55). Consequently I will leave this topic by remarking that the innovator of military gadgetry does have a high likelihood of gaining surprise. Thus he is well advised–and here I merely repeat the standard admonition–to be prepared to exploit his temporary advantage and not dissipate such potentially valuable innovation in real but minor combat trials whose results would thereby he shared with the enemy.66

Although this particular five-fold set of varieties or types of surprise is not found in the military literature, the separate parts do appear in one or another form. Thus “intention” is commonly mentioned although not, as far as I am aware, explicitly identified as a type of surprise. The most common single type of surprise mentioned is “place.” Indeed, many military authorities imply that it is the only one.67 The next most often met type is surprise of “time.” In- deed, many writers mention it together with “place.” A third type of surprise mentioned–almost always together with “place” and “time”–is “strength.” As noted above, some writers also explicitly discuss a special category of “tech- nical” or “technological surprise.”68 Finally, a very few writers recognize that innovations in tactics can yield surprise, although they do not seem quite able to fit that category into a general typology.69 This simple list exhausts the thinking of the standard authors on war.

My theory of stratagem is a decision-making model of alternative choices.70

However, the argument–deriving as it did from Liddell Hart’s theory of “alterna- tive objectives”–drew its examples exclusively from that set of choices involving objectives or goals. All these goals were either specific terrain features (positions or regions) or specific objects (armies or localities) on the terrain. Thus, all the types of objectives were to be attained by moving units of military force over terrain from one area to another. Consequently, the theory may have seemed limited to the “where” or “place” of surprise. However, the model of alternative choices applies equally to the other modes of surprise, specifically to the set of types described in this section: intention, time, place, strength, and style. As all these five modes can be presented as real alternative choices before the pro- tagonist and as alternative choices as perceived by his victim, the protagonist can apply the theory of stratagem to each or all of them in order to enhance his chances for surprise.

The most common mode in which surprise appears is place (or direction), being present in 72% of all instances of surprise studied. Place is closely followed

66This admonition was put forward in most explicit general terms–perhaps for the first time–on 28 December 1914 by Colonel Hankey, the Secretary of the British War Council, in a Memorandum to that high body. Although this particular paper was one of the milestones in the introduction of the tank, smoke, and other devices, his caveat about wasting their surprise effect by field tryouts was ignored. Colonel (later Lord) Hankey was–with Churchill–one of the great technological innovators of both world wars. See Hankey, I (61), 228-231, 244-250.

67Even Samuel Eliot Morison has sometimes slipped on this point. 68For example, Erfurth (38/43), 191-194. 69Although I now find that Miksche (55), 36-37, had anticipated me by a passing reference

in which he gives the four-fold typology of surprise of time (“moment”), place (“direction”), “strength,” and “new weapon or new tactical form.”

70Chapter IV.

114 CHAPTER 5. THE RESEARCH STUDY

Mode of “Strategic” Cases “Tactical” Examples Total Surprise No. Percent No. Percent No. Percent

Place 47 74.6% 31 68.9% 78 72.2% Time 46 73.0% 25 55.6% 71 65.7% Strength 38 60.3% 24 53.3% 62 57.4% Intention 29 46.0% 7 15.6% 36 33.5% Style 16 25.4% 12 26.7% 28 25.9%

Totals: 63 45 108

Table 5.29: Distribution of Mode of Surprise Among Strategic and Tactical Examples.

by time (66%) and strength (57%), trailed by intention (33%), and ended by style, which was present in only 25% of all instances of surprise. As Table 5.29 shows, this trend holds strongly for both “strategic” and “tactical” level oper- ations. The one exception is that intention is better concealed at the strategic level.

Intensity is a second dimension of surprise. There are at least two ways by which the intensity of surprise can be measured. One measure is the number of modes in which surprise was attained out of the five possible modes defined above.71 The other measure of the intensity of surprise is the degree of surprise achieved for any one of the five modes–or for some index or “profile” that sums these five separate surprises. (One can, of course, also test various overall indices that combine both the numbers and intensities of the five modes.)

Although the second measure of intensity is the more realistic one, data on the first is far more readily available. Common sense is sufficient to suggest the hypothesis that the greater the intensity of surprise (by either of these two definitions) the greater will be the effects of surprise, such as more favorable casualty ratios, greater chance of achieving victory, etc. However, the first measure, the one that measures intensity by the sheer number of modes, is directly related to the theory of stratagem. Consequently, the hypothesis that there is a direct relationship between the intensity and effects of surprise can be derived from the theory, which is based on the frequency of alternatives. I will now show that this hypothesis is quite consistent with the data.

Table 5.30 shows the frequencies (number of instances) at each of the five possible levels of intensity of surprise (as measured by the number of modes of surprise inflicted).

In general, we see that the obvious is confirmed: the more intense instances of surprise are rarer. However, there is one striking exception. Surprise very rarely appears in only one of its modes. Inspection of Analytical Lists A and B shows that this is because surprise of place usually coincides with surprise in time or strength.

This general pattern also applies for both “strategic” and “tactical” oper- ations, although it is more or less pronounced at the “tactical” level. While

71As noted by, at least, Miksche (55), 37, who, although he is one of the very few writers to make the point explicit, calls it “a hackneyed truism.”

5.7. VARIETIES AND INTENSITIES OF DECEPTION 115

Frequency Intensity of Surprise “Strategic Cases” “Tactical Examples” Total

1 7 9 16 2 22 20 42 3 18 11 29 4 12 4 16 5 4 1 5

Totals: 63 45 108

Table 5.30: Frequencies vs. Level of Intensity of Surprise

0 1 2 3 4 5 Outcome No./% No./% No./% No./% No./% No./% Defeat (coded D) 35/59.3 4/25.0 5/11.9 0/0.0 1/6.2 0/0.0 Well below expectations (V-) 14/23.7 2/12.5 9/21.4 6/20.7 5/31.3 1/20.0

About as intended (V) 9/15.3 6/37.5 21/50.0 15/51.7 4/25.0 1/20.0

Well above expectations (V+) 1/1.7 4/25.0 7/16.7 8/27.6 6/37.5 3/60.0

Totals: 59/100.0 16/100.0 42/100.0 29/100.0 16/100.0 5/100.0

Table 5.31: Outcome of Operation By Intensity (Number Types) of Surprise

this result could imply that surprise is more difficult to obtain at the tactical level (a very dubious hypothesis as we shall see), consideration of the history of the doctrines of and organizations for deception and surprise suggests a differ- ent cause–namely, that the art of surprise is not as commonly practiced at the tactical level.

We would expect that the degree of success in a military operation would vary directly with the intensity of the initial surprise. That is, the chances of the commander achieving his intended objectives or his pre-battle estimates of the outcome are markedly enhanced by the intensity of surprise inflicted by his initial stroke. The empirical evidence supports this, as shown in Table 5.31. While most of us would consider this a truism, I doubt that any would have expected the degree to which it is true. The trend far exceeded my expectations both in its steepness and in its consistency as analyzed across both dimensions. To cite only the extremes: Out of 59 battles fought without any initial surprise, only 2% substantially exceeded its general’s expectations while 60% ended in abject failure. Conversely, out of 50 battles where surprise was intense (rated 3 or more on a 0-to-5 scale), 34% far exceeded their objectives and only 2% ended in defeat.

Deception also can be typologized in all its variegated forms and measured by its intensity. I have tentatively adopted a rather rough-and-ready typology (Table 5.32). Moreover, the data is of poor quality, because there are very many cases where I know of only some of the specific ruses or types of ruses used. For

116 CHAPTER 5. THE RESEARCH STUDY

Type of Deception “Strategic” “Tactical” Total

Feints (deployments simulating an imminent attack) 32 17 49

Dissimulative camouflage (concealed installations) 29 18 47

Simulative camouflage (dummy installations) 32 12 44

Demonstrations (diversionary attacks) 26 11 37

Rumors (deliberately planted) 25 5 30 Radio (deceptive traffic,

excluding psywar) 22 7 29 Press leaks (including

public announcements) 20 0 20 Negotiations (diplomatic) 13 1 14 Fake documents 6 2 8 Other 12 6 18

Totals: 63 45 108

Table 5.32: Decomposition of Deceptions in a Rough Typology.

example, I have certainly missed many cases in World War II where field radio was used to simulate and dissimulate deployments. Nevertheless, the data and categories are probably adequate for an initial look.

No comment on Table 5.32 seems needed, except perhaps to note that those types of deception more appropriate to larger scale operations–fake press leaks, diplomatic deception, and even rumors–are indeed more common at the “strate- gic” level.

A sharper look can be taken at the number of feints (deployments simulating a threat to a place) and demonstrations (real but diversionary attacks). This is justified for two reasons. First, the quality of data on both categories is quite good. Secondly, as both categories comprise major means of threatening “alternative objectives,” they relate directly to the theory of stratagem. For example, the following hypothesis72 may be tested. The larger the number of geographical points or directions threatened (i.e., approximately the sum of the feints and demonstrations), the greater the dispersal of the victim’s forces (as roughly measured by unfavorability of his force ratio to that of the deceiver at the point and time of the real attack). Table 5.33 results.

The hypothesis is generally confirmed. More feints are associated with more favorable strength ratios, as Table 5.33 shows.73 However, the detailed analysis showed that there was only a slight marginal advantage gained by increasing the number of feints and demonstrations above two. The reader is cautioned that my choice of using strength ratios at the point of attack (or as near to

72Suggested by William R. Harris, September 1968. 73The actual ratio intervals are 0.01-0.3; 0.4-0.6; 0.7-0.9; and 1.0-14.0, where the number of

units in the victim’s force are divided by the number of opposing units.

5.7. VARIETIES AND INTENSITIES OF DECEPTION 117

Number of Feints and Demonstrations 0 1 2-7

Strength Ratios No. % No. % No. %

At least 3-to-1 7 16.3 9 32.1 14 45.1 About 2-to-1 9 20.9 6 21.5 6 19.4 More than 1-to-1 12 27.9 3 10.7 5 16.1 Less than 1-to-1 15 34.9 10 35.7 6 19.4

Totals: 43 100.0 28 100.0 31 100.0

Table 5.33: Feints and Demonstrations vs. Strength Ratios.

Number of Places Threatened Degree of 1 2 3-4 5-8 Surprise at Place No. % No. % No. % No. %

None (0) 23 47.0 8 26.7 4 15.4 0 0.0 Some (S-,S) 18 36.7 15 50.0 11 42.3 2 28.6 Much (S+) 8 16.3 7 23.3 11 42.3 5 71.4

Totals: 49 100.0 30 100.0 26 100.0 7 100.0

Table 5.34: Relationship Between Surprise and Number of Places Threatened

that sector of the front as I could get figures comparable for both sides) is an unsatisfactory measure of dispersal of the enemy forces. A better measure would be to compare the proportions of their whole forces, which each had deployed at the threatened portion of the front, a statistic that I did not collect.74

Similarly, we can pose the hypothesis that the larger the number of geo- graphical points threatened, the greater the chance of surprise (and the greater the degree of that surprise) at the point of the real attack. Again, the data is consistent with the hypothesis. Indeed, the trend is entirely consistent even in the full interval table of which Table 5.34 is only a summary.

The above analyses add to the impression that surprise does confer many great rewards to the side that inflicts it. Thus, any steps that a commander can take that will increase his chances of surprising his enemy should receive his attention. The central assumption of the theory of stratagem is that deception is the most effective means of doing this. So far, I have done no more than show the coincidence of deception and surprise and to argue from anecdote that the former causes the latter. Quantitative analyses of the type used here cannot prove cause. (After all, generals who think deceptively could all belong to some special type with a “winning” personality.) However, they can account for the anomalies (as, for example, when we found that the avoidance of Fridays as a day for launching operations was a World War I phenomenon). Moreover, our confidence in the likelihood of a causative relationship of deception to surprise

74The ideal measure would be the proportionate changes in force at the threatened point during the period of feints and demonstrations. However, such detailed information is seldom available for any given operation.

118 CHAPTER 5. THE RESEARCH STUDY

Intensity Intensity of Deception of Surprise 0 1 2 3 4 5 6 7 8 9 10 Total

0 1 1 3 1 6 1 6 5 1 1 2 1 16 2 7 7 7 12 6 1 2 42 3 4 3 9 6 5 1 1 29 4 1 2 2 3 2 1 4 1 16 5 2 2 1 5

Total: 20 18 20 25 15 5 2 7 2 0 0 114

Table 5.35: Relationship Between Intensity of Surprise and Deception.

Intensity of Deception Intensity 0 1-4 5-8

of Surprise No. % No. % No. %

0 53* 72.6 5 6.4 1 6.3 1 6 8.2 9 11.6 1 6.3 2 7 9.6 32 41.0 3 18.7 3 4 5.5 23 29.4 2 12.5

4-5 3 4.1 9 11.6 9 56.2

Totals: 73 100.0 78 100.0 16 100.0

Table 5.36: General Trends in Relationship Between Intensity of Surprise and De- ception. *Note: This cell is represented by all 53 Type C Examples. While it is not possible to say how representative the figure may be, the bias in selecting Type C was to err on the low side.

is increased as various corollary hypotheses are also tested and verified. The immediately preceding table was an important example in that it demonstrated the predicted relationship between one specific form of deception and one specific form of surprise.

The relationship between the intensities of surprise and deception are shown in Table 5.35.

The general trend can be seen more clearly in Table 5.36, which glosses the minor trends and perturbations.