-Top-Secret
Scientific and Technical Intelligence Report
Capability of tbe Soviets to Train Marine Mammalsilitary Operational System
/
CAPABILITY OF THE SOVIETS TO TRAIN MARINE MAMMALSILITARY OPERATIONAL SYSTEM
Project Officer
6
CENTRAL INTELLIGENCE DIRECTORATE FOR SCIENCE AND TECHNOLOGY OFFICE OF SCIENTIFIC INTELLIGENCE
8
Capability of the Soviets to Train Marine Mammalsilitary Operational System
Project Officer
PRfiCIS
Within the next year, the Soviets could train marine mammals to be used In military operational systems such as diver assistance and equipment recovery In the Black Sea, The nuct steps, possibleears, might be training of the animals for more sophisticated tasks such as placement of packages on ships and ship protection, as well as for use in areas outside the Black Sea. The training achievements would be contingent on continuing Soviet success intable of healthy animals and overcoming past training difficulties.
The Sovieis reportedly experienced difficulties In their early marine mammal training programs, but many of these difficulties appear to have been overcome. Useful experience has been gained in capturing and maintaining the animals and in providing veterinary care which are important prerequisites to behavioral training.w facilities to support the marine mammal program have been constructed. Recent evidence indicatesoviet capability to train the animals for use in the open-ocean, at least in limited military and Intelligence-gathering roles, may now be available. ,
7 K
PREFACE
Indicates that the Soviet Union may be attempting
to train mai.ne mammals for military purposes. This report briefly reviews the prerequisites foruccessful training program, summarizes sone of the major difficulties experienced during training, and reviews Soviet progress In surmounting such difficulties In order to develop operational programs. The operant conditioning and classical conditioning used to train animals are discussed In an Appendix,
Marine mammals can perform operational tasks which are well beyondrained frogman'san with swim flippers canpeednots for several minutes In quietolphin can cruisenots for several hours and sprintpeed of IS tonots for several minutes.eet toree swimming diver hasinutes of working time if he wishes to return to the surface directly. According to US Navy decompression tables, he cannot make another dive forours. Using revolutionary equipment developed Iniver can descendeet for upours and require no decompression. Lack of mobility, the inefficiency of the mechanical hands, and Inability of the diver to operate In murky water, however, places significant limitations on this eiperlmental system. Dolphins, on the other hand, can diveeet, stay for several minutes, make repeated dives and ascend quickly without experiencing decompression problems. Whales are capable of diving to depths of ateet und can stay submerged for about an hour. Man's ability to localize utiditory signals is poor In water; in murky water his vision is restricted severely. By contrast, some marine mammals which use sonar (both active and passive) can locate both sound-emitting and silent objects as small as one and one-half Inches long.'
The abilities of marine mammals can be "shaped" for operational purposes. For example, marine mammals have been trained to detect and intercept swimmers, tc locate objects, to place Instrument packages on moving or stationary targets, and to carry tools, lines, and objects from the surface to divers and between submerged divers.
This paper Is concerned with mammals from the Order Cetacta (carnivorous, wholly uquotlc mammals Including dolphins andhe main focus Is on Tuntopi tnmcaiiis, the bottlenose dolphin, which Is the only marine species known to be used by the Soviets for operational training.
Information was obtainedliterature, and
scientific interchanges with the uasn.by the Office of Scientific Intelligence and coordinated within CIA. The cutoff date for Infcimatlon Is
BLANK PAGE
CONTENTS
Pope
1
SUMMARY AND1
2
Genesis of Soviet Marine Momma!
Marine Mammal Research and Training
Capture and
Training Personnel
Training
APPENDIX
Page
Trainingvs. Classical9
CAPABILITY OF THE SOVIETS TO TRAIN MARINE MAMMALSILITARY OPERATIONAL SYSTEM
PROBLEM
To cvaluotc Ihc Soviet ability to trnln marine mammal* for use In military missions In the open ocean.
SUMMARY AND CONCLUSIONS
knowledge of US successes and recognitionarine mammal program gave Impetus to the Soviet efforts in this field. The Soviets are developing operational systems based on an analysis of the unique capabilities of dolphins and their own needs. They are also assessing and replicating US systems while possibly developing counfermeasures to certain US systems. The Soviets have written of the threat posed by the US effort, stressing that the most significant concern is the effect of mammals on submarine forces. Tor example, animals could be trained toagnetic packageubmarine which could record information,racking signal or eiplode.
At present, training for operational programs is done onlyhe Black Sea area. Such training probably has been intended toimited capability such as diver assistance or equipment recovery.
The Soviet program initially faltered because of difficulties in capturing and transporting dolphins without harmingack of expertise in maintaining dolphins in captivity and little or no expertise In dolphin veterinary medicine. Since sick or injured animals do not respond torerequisite for opera ial training was to overcome the above difficulties. Early difficulties also resulted Iron using Inexperienced navy conscript! to train the dolphins.
The Soviet! have several dolphin research and training facilities which are at least equal to those in the US in size and availability of holding space. In addition, the Soviets appear to be resolving some of the problems Involved In maintaining healthy animals over the period of years necessary to develop opera tlonal programs.
Recent assessment of Soviet training skills shows some Improvements. Evaluation (if the abilities of"
the black Sea also may lead to the establishment of bases on the Soviet east coast, where temperature and other operational conditions are appropriate. The most likely place foracility would be near Vladivostok.
The early stages of the Soviet program were hampered by several problems. Existing field stations had to lie remodeled and new facilities had to be built before research and animal training could commence.
Acknowledging the limitations of generalization ' 'he Soviets now have
leust some good trainers, some of wham may not be In the Navy. There Is still some question whether the SovIe> Navy has used this expertise to train its dolphins, to Improve the capabllltlei of Its own truliiers. or bolh.
Top^Sacrel
toft iSatrr'
DISCUSSION
OF SOVIET MARINE MAMMAL ;. RESEARCH
The mafor Initial stimuli to the Soviet marine mammal research program probably were the reports on or allusions to the efforts and accomplishments of the USrief discussion of several publications which the Soviets are reported to have studied may provide an insight Into the nature of the threat which the Soviets perceived.
4 Dr. J. C. Lilly's book. Man and Dolphin, was translated Into the Russianhe main theme of the book waslphlns were at least as Intelligent as man and that twj-way communication between the species would occur within two decades.
An excerpt from the book which addresses the possible military use of dolphins follows:
"CrtaCToni might be helpful In hunting and retrying nu* cnnei, utellltei, mlwllrs. and similar things that men IrHit on dropplnr. Into Ihe oc*sn. They might be willing lofor mines, torrxdoes, submarines and nthwartifacts, tod*scouting and patrol duly tor submarinei or surface ships, and they might carry their pern agonist activities to the point where they can beound ha.bon bi underwater demolition-team
The Soviets were aware of the US Navy's Interest in marine mammals. In his book, Dr. Lillythe help provided by the Office of Naval Research. Department of Defense. Thismay have been Interpreted by the Soviets as an Indication that the US Navy friended to use marine mammals to develop new weapons or to function as integral portions of weapon systems.
Such an impression also could hav; been fostered by an earlier book in which Harrison Matthews had lamented that, "now It seems that some people ate proposing to prostitute their biological work on the Cetacca and involve the animals in human international strife by training them as underwater watchdogs to guard naval installations from frogmen or to act as unmanned submarines."1
rticles appeared in several US newspapers revealing the "fact" that the navy was training Kamikazeheme which has appeared subsequently. As recentlyhe London Newt reported an unconfirmed incident In whicholphinonitoring device on the hulloviet nuclear submarineoreign harbor and collected it later so that the type of fuel the submarine was using could be discovered."4
Reportedly, the task of analyzing US unclassified literature and classified marine mammal research was assigned by the Soviets to the Office of New Technology of the General Staff of the Ministry of Defense. The head of the Bionics Department of the Office of New Technology was reported to have shownob candidate confidential reports In the Russian language concerning US Navy marine mammalhe Soviets apparently had an effective collection effort, as US marine mammaloglsts visiting their counterparts In the USSR were surprised by the Soviets' familiarity even with obscure publications.'
From the analyses they performed, the Soviets evidently concluded that the US Navy hud trained dolphins to perform military tasks successfully.5 article In Knunaya Zvezda, Lttcnctskly discussed the use ofy the US Navy and itst is certain that the Soviets regarded marine mammalsotentiul military intelligence threat; they may have realized the potential use of the animals to gather data concerning submarines which, to seme extent, would compromise the effectiveness of this important strategic force.
MARINE MAMMAL RESEARCH AND TRAINING
The Soviet State Committee for Science and Technology reportedly approved an Academy of Sciences proposal to make dolphinovernmental responsibilityeports on Soviet marine mammal research have Increasedhere are three major areas of the Soviet program as Indicated by the number ond content of the research articles produced. US marine mammaloghts hoveecent Soviet articles. Some were classified In more than one category. The three categories of interest to this paper are us follows:. hydrodynamles-
Tisp^aetV
..litcdnd bchuvloral aspects.alning artlclci dealt with miscellaneous
took*.
Facilities
The Soviet marine mammal research hasat various institutes and field stations.provide indirect support to fieldperform applied research andAll of the Identified field stutlons are onSea. The field stations at Sukhumi,Karadag. and PUsunda-Alakhadiyas being Involved mainly InAlthough acoustics, physiology,medical research also Is pursuedstudies are central to tliethere. In
suitable for training marine mammals for operational tut*.
Construction of thebegun
t Is now ;alrly extensive and Incorporates several large researchee^neaer hydrodynamic ru.iwuy. and large enclosed pens with both holding and trainingndnasmuch as [and supportieratlonal training need not consist of more than an Ice house to store fishhed to store equipment, the research buildings probably are largely devoted to hydrodynamics and acoustic research.
] may Indicate that
o achieve this goal, ihe Soviets needed lo remodel old facilities, or build new facilities, and to leant the Intricacies of capture, maintenance, and training."
! iCapture and Maintenance
Regular fishing fleets supply animals to the Soviet research stations. The fishermen Initially were not trained In the procedures necessary to ensure Ihe well-being of these anlmali, Even with researchers present to supervise the catch, many animals died during capture, transport, or soon after their arrival at the
research facilities 1
ofursiops caugVir, oniy two survivca.eport In
ndicated that of IS animals caught, only two were aliveincenimals were Injured or traumatized during handling, considerable time had to be spent nursing the anlmab backhealth before effective training could begin."
The Soviets alsoxperienced numerous problems Ininter dolphins In open-sea pens
that two tuniops
on the Black Sea.
were wintered for the firstReportedly, one animal was maintainedean In captivity and anotherearsilch Indicate at least some capability was galnesr in winteringevertheless, as recently aster, the Karadag facility lost all but one of Its
"TSp^cra^r
Tofcjatjr'-*
even thoughJLalso hod heatedot clear whether the re maintained In npcn-seu pens, or were broughtheltered enclosure during the winter months. Some of | nimalse wintered at the Batumi fa.-illty.
T'irslops coniume approximatelyounds ofdiy In temperate weather; In cold weatherIncreases dramatically. Feedingogistics problem.of fish, and the right kind of fish mustavailable, Animals develop strongduring captivity and generally will reactIn the; diet by refusing to eat. Theat times did not provide the pruper typeor follow the necessary hygienico the logistics difficulties infresh fish of the proper species Is, ofuse frozen fish, but this also has presentedthe Soviets. Dolphins which are fed frozenthawed fish develop stomachare frequently fatal.
caretakers feeding partially-nSvvVd lUh to tho dolphins.11
There Is as yet no Indication that the Soviets are developing any type of artificial food. An artificial food could eliminate the problems caused by inadequate thawing of frozen fish. Research lu this area has been conducted by the US Navy,1
An additional maintenance problem for dolphin researchers Is the quality of veterinaryasic procedures such as prophylactic use of antibiotics, reportedly were not employed by the Soviets In spite of the fact that captured dolphins arc quite susceptible tohe significance if inadequate veterinary core Is that sick animals do not respond to training. In general, veterinary care has been one of the major problems of the Soviet program. The Soviets now appear to be Improving their veterinary care and resolvingof the problems Involved with maintaining healthy animals.19
Training Personnel
7 Callna Shurepova became head of the Snvkt applied researchaslef of underwater sportormer scuba diving champion There Is no Indication, however, that she initially possessed any particular training skills or knowledge of sea mammals. Sabltovormer
Soviet Navy officer andcubu diver, acted as her assistant.ndividuals were In chargenit of nuvyof whom weie conscriptedperformed the everyday care and feeding as well a* the training of the animals.'*
The use of conscripted personnel for animal care and training maylnderance If the results of US experience are applicable to the Soviet case. Conscripted personnel. In general, lack the motivation, perseverance and attention to detail required to train animals. The reports of Initial failures by thedd credence to the hypothesis that conscripted personnel are not effective animal trainers.
Because of the difficulties encountered by the Soviets In training dolphins. Anna V.ircus trainer who had experience training seals, was uppolntedechnical advisor. Her father, Vladmlr Durov, reportedly had developed before World Warystem employing seals and sea Hons to detect and neutralize mines andystem Is probably the one. Skinner'ychologlst who founded operant conditioning and trained animals for the US military) alluded to but lacked supporting evidence.'* It appears that Anna was selected forexpertise and also because of her father's previous work in developing marine mammal systems.
Conversations with Durova4 Indicated that she felt that no reliable results had beenShe felt that the large number of animals receiver! for training was out of proportion lo the small number of skilled handlers, The trainers annually were faced with an InPux of new nnimoli to replace the dolphins that hod died during theiy animals also were lost In several programs which attempted to train onimals to return to their pens at the base aftet being released in theertain of Durova's comments were critical of the trainers' abilitta.'* Thus, there is still some question whether the Soviet Navy has used the civilian expertise to train either the trainers or the dolphins, although recent ev'dence indicates some accomplishments in the navy's training
Training Te.-hnlques
The US had seven! advantages compared to the USSR when It began Us marine mammal operat'unal program. In the US, experts were available who had worked with marln" mammals in various oceanariums.
The Soviet Union did not open un occunurlumnd there wui no pool of expertsith the training nf dolphin*
There known for having excellent circus trulnen. but none was known to have worked with dolphin? Also. Ihe ability of circus trilnen und the effectivt nets of their training techniques sometimes vary greatly from those needed for training animals to perform specific tasks. Circus trainen tend to build acts Incoeputatlng what the animal does rather than training the animal to perform other specific tasks. Circus trainen tend to be unaware of what actually Is effective In their mining technique and have developed many supentlllous Ineffective training behavlon to which they mlttakenly attribute much of their success. Much of the circus trainen' effectiveness can be attributed to the operant conditioning procedures he unwittingly uses. The traditional Soviet clrcu* training method also relies heavily on punishment and fear, both of which have been shown to be ineffective In training dolphins.'
The US may have enjoyed an Initial advantageprogram In 'nat operant conditioning, which Isbased reliable method of trainingdeveloped In the US. The Sovietsoperant conditioning because of Itswith Pivlovian classicaldolphin research using operantto lack many of the refinement! seenresearch. The vast amount of USand using operant conditioning IsIn Soviet open(An
appendix discusses operant venus classicalas methods of training anlmab.)
A brief examination of some problems encountered by the US program Indicates the complexity and organizational problems involved In working with marine mammals. Similar difficulties may have been or may continue to be encountered by the Soviets. Some of the difficulties encountered In the USare illustrated In the followingbuzzerolphin attracted sharks and the dnl-poln turned out lo sea and Ignored the recallights and noises In the operational setting did not match conditions existing during training and thus leday delayolphin performedrained dolphin Inexplicably felled to perform but upon subsequent examination. It was revealed that the target burner had failed to operate.4
ined sea lions exhibited erratic bchuvlor In recovering objects,ubsequent study revealed thut the fog und tbe uccumpunylng durkness ut the recovery site Interfered wilh the onimabdditionally, tlie pattern painted on the object differed from the pattern used duringhus the fidelity between the training ond the operational setting und stimuli, communication between the truinrn and equipment suppliers, and knowledge of the limitations of the anlmilt ,nd eoulpment Ktvn hindered the US program, und liieseproblems could be expected to be encountered by any bureaucratic organization such as the Soviet Navy.
From the Soviet experiments reported In the open literature. It Is Impossible lo evaluate the competency of Soviet animal trainen. their training techniques or the efficiency which they may have dealt with problems such oi those enumerated above. Censon reportedly remove most of the methodological sectionfrom technical papen before they are accepted for publication, and It. therefore. Is unavailable for analysis. The methodological information is customarily conveyed through lectures andIt Is Impossible, therefore, to determine the validity of even the limited results which are reported.
It Is likely that, at least Initially, the Soviets did not have the expertise In using complex conditioning techniques. Because some operant techniques have been adopted by Pavlovlan-oriented scientists, it appean that the Soviets nowudimentary familiarity with the empirically based operant method of training animals. Much of the Soviet dolphin research requires behavioral responses similar to those demanded In UShese studies necessitate the use of some form of operant conditioning. Although the Soviets use some operant techniques, they continue to use Puvlovlan terminology and concepts Ir discussing their experiments.
Top Stcrfi
was given ihe task nf trainingwo-choice discrimination taskto assess the sonar capabilities of theunlmals had not been trained or used InThe training took place In anfrom US (miners In order lo
much as possible.
During the training.not verbalize or
acknowledge operant conditioning techniques, but he used them effectively. Especially noteworthy was his effective usebridgingignal to the animal that It has performed correctly and willcwurd. The bridging stimulus must be presented immediately after the correct behavior Is performed und is used to reinforce the animal when provision of an immediate reward Is inconvenient or undesirable. Other ope.ant techniques such as "shaping" und "time-outs" also were apparent In his training regime.
eeks, the animals were stationing, gale training, and performing an initial discrimination. Considering his evident skillrainer, and the recently reported accomplishments of Soviet marine mnmmulhe Soviets now have atimited training expertise toarine mummal operational system,I
Unlike US training, Soviet taming and training emphasize physical contact with the animal. The standard Soviet taming techniques Involvetraining animabat food from the hand of theermit its body to be twichedllow the trainer to touch, grasp, or holdngagr In uctlve play with the trainer;llow forceful manipulation of its body. Rough gumes are viewed by Ihe Sovietspecial way of establishing Intimate contact. Reports of trainen beating on animals' rostrums are Included In descriptions ofecent report, the civilian trainen at Moscow Slut* University found thai the uv ol on infant dolphin greatly facili'ated successful training of these five stages of behavior. Since the Soviet Navy mulntaliH its unlmals In bisexuul groups. Il alio may be lurcrcstcd In using Infant dolphinsrogram."
Along with, or perhaps because of. the Intimate contact with the animals, the Soviets stress that taming jnd training are done best by one man. While acknowledging the desirability of huvlng the animal respond to several trainen rather than one specific trainer, the Increase in effectiveness, as Judged by the Soviets, appean to warrant the one-to-oneA team concept of training Is more appropriate for an operational system. If several men were able to handle the animals, mission* would not be delayedundler was sick, on varation, or left the service. The Initial lack of successful training may have convinced the Soviets that they could nol afford the udded utility provided by the use of multiple trainen.
The Soviets appear to be aware that. In addition to fish, effective reinforcements for the animals can be obtuliird through interaction with men, play objects,arge swimminghe Soviets do not appear to apply the rcinforcenependent on the performance of the animal. Provision of non-food relnforcer* is not made contingent on the animals behavior; thus, much of the usefulness of the relnforcer Is lost.
Information needed for irulnlng marine mammals Is widely available In the US open literature. The Soviets are aware of this literature. Nevertheless, awareness does not necessarily coincide with application. In several studies, the Soviets have reported an Inability to trainn other studies. It is apparent that the five major steps in operant training have not been
jor steps in the application of operant techniques In order to maximize results requires: (I) precise detailed description of behaviorallanning of steps which allow shaping of animals so that ihry perform tasksccurate recording of behavior with concurrent analysis In order that animal's progrct. or lack of It, can behange from continuous reinforcement to variable Intermittent reinforcement schedules, and (S) variation of tasks in order to maintain the animal's motivation.
jret
50
An apparent contradiction exists between the results of the evaluation of the skills displayed by('
his evaluation of the Soviet program. At one tli
therefore, may have
a
'ashotographilot whale rklng deviceS torpedo. He stated that It could take thr0 yean to get to thatet our evaluation Indicates that Soviet civilian trainers and consultants to the military program possess some ability In training. Part of this dlscrepency may be attributed to the Soviets' Inability to capture andhole which, until recent years, was
considered Impossible
been referring tomot whale rather than training marine mammals In general.
IMPLICATIONS
The presence of the lubmersibles at andof training involving physical contact withsupport the hypothesis that one of thegoali Is tolrtrThe Soviets also have attempted toanimals to respond to radio controls and topackages."
possible portable ect. the Soviets arc
hliunats. ii mrsimcrpieiauuiiiscut alsoapability to transfer and maintain the animals at temporary locations. The Soviets are fully capable of relocating the animals and have transported marine mammals by plane, boat, and truck. Portable habitats would permit maintaining the animals for Indefinitely long periods at locations remote from the main facility, as would be required for ctrtaln operational activities.
The Soviets most likely will develop their own operational programs, will continue to assess the US marine .iiammal program, and probably will attempt to replicate US successes in order to develop countermeasures to certain programs. The Soviets first would attempt to use the animals In their home waters In tasks such as diver assistance or object recove*y. Once the Soviets are successful with these tasks, they may expand their program beyond the limits of the Black Sea, The Soviets would expand their training facilities to other waten under their control, such as on the east coast, The Vladivostok area appears tot suitedarine mammal facility. Adequate naval support is present In thisase In Vladivostok could serveome base for operations In areas such as the North Pacific, Sea of Okhotsk, and the Sea of Japan.
After Initial successes, the Soviets may attempt to train their animals for more difficult tasks such as guarding ships in International waters or attaching packages to foreign vessels In order to record Information, toracking signal, or to damage the vessel.
The Soviets could within the next year train animals to perform reliably In the Black Sea such tas& as object recovery or diver assistance.ears they could begin using the animals In other ocean an- as and have animals trained to perform more difficult tasks.
Top>*atret
Top
Trainingvi Classical Conditioning
of the conditioning paradigms wilt clarify the differences and similarities between operant conditioning mainly used In US research and classical conditioning which ptHomlnates in Ihe USSR. The standard procedure of operant conditioning involves giving the organism which is beingeinforcement Immediately after It performs correctly. The Immediacy of the reward and the performance of behavior are the key elements in conditioning. By rewarding successfully closer approximationsesired behavior an animal can be "shaped" toew behavior. Thus operant conditioning depends on the animal emitting some aspects of desired behavior.-
In classical conditioning, onestimulus (CSMs paired repeatedlyecond, unconditioned stimulus (UCS) until the CS by itself elicits the response normally elicited by the UCS. The key elements in classical conditioning are that the CS Immediately precede the UCS and that the interval between CS and UCS presentation be as short as possible. In classical conditioning, the sequence of events Is not dependent on the animal's behavior.
Important propertlw of the two methods can be combined. Operant conditioning requires that the reinforcement be contingent on the occurrence of the desired response or behavior. In classicalew stimulus ultimately takes on the propertiesatural, unconditioned stimulus and elicits the same response. In advanced conditioning trials, components of both classical and operant conditioninglassically conditioned stimulus may be usedsecondary relnforcer" In an operant conditioning trial to Inform the animal when it has performed correctly. This relnforcer can then be used as the stimulus for training different responses. For example, the usehistleecondary relnforcer for food makes it easier to "reward" the hungry animal Immediately after it has performed the correct behavior and to encourage It to perform an additional desired action.
Top
TaVsecret
Original document.
Comment about this article, ask questions, or add new information about this topic: