英：transference; 法：transfert; 德：Ubertragung
移必“转移”这个术语首度出现于弗洛伊德的著作，仅仅是表示情感 (affect)从一个观念 (idea)转移至另一观念的移置 (displace-ment)的另一名称 (见：Freud, 1900a: SEV, 562)。然而，该术语后来渐渐开始指涉病人与分析家在治疗中所发展起来的关系。这很快就变成了此一术语的核心意义，而且现今的精神分析理论通常也都是在这个意义上来理解它的。
用一个特定术语来表示病人与分析家的关系，其依据即在于此种关系的特殊性质。在1882年布洛伊尔对于安娜·欧的治疗中，弗洛伊德首次遭到了病人对医生做出的那些强烈情感反应的冲击，他认为这是由于病人把那些无意识的观念转移到医生身上的缘故 (Freud, 1895d)。随着弗洛伊德发展出精神分析的方法，他起初只是把转移看作一种阻抗 (RESISTANCE), 它会阻碍对于那些被压抑的记忆的召回，是必须遭到“摧毁”的一种对于治疗的障碍 (Freud, 1905e: SEVI, 116)。然而，渐渐地，他修改了此种见解，而开始把转移也同样看作有助于治疗进展的一个积极因素。转移的积极价值存在于这样的一个事实，即它提供了一种方式，在当下与分析家的现时关系中来面对分析者的历史，在他与分析家建立关系的方式中，分析者不可避免地会重复与其他人物的那些早期关系（尤其是与父母的那些早期关系）。转移的这一悖论性本质，即同时作为治疗的障碍与驱使治疗前进的动力，或许便有助于说明在当今的精神分析理论中为何会有这么多不同且对立的有关转移的见解。
拉康关于转移的思考经历了几个阶段。他详细处理这个主题的第一部作品是《关于转移的发言》(Lacan, 1951), 他在这篇文章中是根据借自黑格尔的辩证法来描述转移的。他批评自我心理学根据情感 (AFFECTS)来定义转移，“转移并不指涉情感的任何神秘特质，即便当它以情绪的外观而被揭示出来的时候，它也仅仅是凭借它被产生的那一辩证时刻而获得意义的”(Ec, 225).
换句话说，拉康认为，尽管转移常常会假借那些特别强烈的情感而表现出来，诸如爱 (LOVE)与恨等，然而它并非是由这些情感所构成的，而是存在于一种主体间关系的结构。转移的这一结构性定义始终是贯穿在拉康其后著作中的一个恒定的主题，他坚持把转移的本质定位于象征界，而非想象界，尽管转移也明显具有一些强大的想象性效果。后来，拉康又评论说，如果说转移往往是在爱的外观下表现出来的，那么它所涉及的也首先是对于知识 (savoir)的爱。
在1953一1954年度的研讨班上，拉康又返回了转移的主题。这一次，他没有根据借自黑格尔的辩证法来构想转移，而是根据借自人类学的交换来对其进行构想 (莫斯、列维-斯特劳斯)。转移内隐于言语的行动，它涉及的是转换言说者与倾听者的某种符号交换：
在其本质上，我们认为有效的转移只不过就是言语的行动而已。每当某人以一种真诚且充分的方式对着另一个人言说的时候，便产生了真正意义上的转移，即象征性的转移发生了某种事情，从而改变了当下的两个存在的本质。
(S1,109)
在接下来一年的研讨班上，拉康又继续详细阐述了转移的象征性本质，他将其视同于强制性重复，即主体的象征性决定因素的坚持 (S2,210-11)。这要与转移的想象性面向区分开来，即爱与侵凌性的情感反应。在转移的象征性面向与想象性面向之间的这一区分中，拉康提供了一种有益的方式来理解转移在精神分析治疗中的悖论性功能，在其象征性的面向 (重复[REPETTTION]), 转移通过揭示出主体历史中的那些能指而有助于治疗的进展，而在其想象性的面向（爱与恨），转移则充当着某种阻抗（见：S4,135; S8,204).
拉康再度着手转移的主题，是在其简单题为“转移”的第八年度的研讨班上 (Lacan, 1960-1)。在此，他运用柏拉图的《会饮篇》来阐明分析者与分析家之间的关系。阿尔喀比亚德把苏格拉底比作装有某种珍贵对象 (希腊语：agalma, 即“神像”)的一个普通盒子，正如阿尔喀比亚德把某种隐藏的珍宝归于苏格拉底那样，所以分析者也会在分析家的身上看到其欲望的对象（见：对象小a[OBJET PETIT A]).
在1964年，拉康把转移的概念链接于他的假设知道的主体 (SUBJECT SUPPOSED TO KNOW)的概念，从此时起，这个概念便始终是拉康的转移观的核心；实际上，正是此种转移观渐渐被看作拉康旨在理论化这一问题的最完整尝试。根据此种观点，转移是把知识归于大他者，假设大他者是一个知道的主体：“一旦假设知道的主体存在于某处…转移便发生了”(S11,232).
尽管转移的存在是精神分析治疗的一项必要条件，然而就其本身而言，它是不充分的，分析家以一种独特的方式来处理转移也是同样必要的。正是这一点把精神分析与暗示 (SUGGESTION)区分开来，虽然两者皆是以转移为基础，但是精神分析不同于暗示，因为分析家拒绝使用转移给他赋予的力量 (见：E, 236).
在精神分析的历史上，从相当早期的时候开始，人们便在区分病人与分析家关系中的那些“适应现实”与“不适应现实”的方面达成了共识。但凡是由“以一种扭曲的方式来感知分析家”所导致的那些病人的反应，统统都会落入后一范畴中。有些分析家用“转移”这一术语来指涉分析者与分析家关系中的所有面向，就此而言，他们便把经过扭曲的“神经症性转移”(neurotic transference)或“转移性神经症”(transference neurosis)区分于“转移中无可非议的部分”(unobjectionable part of the transference)或“治疗联盟”(therapeutic alliance)(诸如爱德华·比布林[Edward Bibring]和伊丽莎白·蔡策尔[Elizabeth Zetzel])。另一些分析家则认为，“转移”这一术语应当仅限于分析者的那些“不现实”(unrealistic)或“非理性”(irrational)的反应（诸如威廉·西尔弗伯格[William Silverberg]和弗朗兹·亚历山大[Franz Alexander])。然而，潜藏在这两种立场之下的共同假定，都在于分析家能够分辨出病人针对他的反应在何时并非是基于他的真实人格，而是基于同其他人的先前关系。分析家之所以被认为拥有此种能力，是因为他理应比病人更好地“适应现实”。由于分析家了解其自身正确的现实感知，他便能够提供那些“转移的解释”(transference interpretations), 也就是说，他能够指出真实情境与病人对此做出反应的非理性方式之间的不一致。据称，此种转移的解释有助于分析者获得深入其自身神经症性转移的“洞见”，从而消除转移或是“肃清”转移。
拉康最尖刻的一些批评，都针对的是此种表现精神分析治疗的方式。这些批评皆是基于以下的几点论证：
(1)适应现实的整个观念都是基于一种天真的经验主义的认识论，涉及诉诸把“现实”作为某种客观且自明的给定的那种没有问题的观念。这全然忽视了精神分析的发现，即自我是经由其自身的“误认”(meconnaissance)来建构现实的。因此，当分析家假定自己比病人更好地“适应现实”的时候，他便无非是在求助于“退回到他自己的自我”，因为这是“他唯一知道的一丁点儿现实”（E, 231)。病人自我中的健康部分，于是便被简单地定义为“像我们那样去思考的部分”(E, 232)。这便使精神分析沦落成了某种形式的暗示，分析家在其中仅仅是“把他自己的现实观念强加”在分析者的身上 (E, 232)。因而，“【分析家】无法以一种真正的方式来维持一个实践，如同人类常有的情况那样，其结果便导致了权力的实施”(E, 226).
(2)凭借解释能够肃清分析者“对于分析家的扭曲感知”，这种思想是一种逻辑上的谬误，因为“转移的解释既是以转移本身为基础的，也是以转移本身为工具的”(S8,206)。换句话说，根本就没有转移的元语言 (METALANGAUGE), 在转移之外没有任何能够让分析家提供某种解释的有利位置，因为他所提供的任何解释都“将会被认为来自转移将其转嫁给的那个人”(E, 231)。因而，当转移本身恰恰规定着分析者对于那一解释的接受的时候，这种主张通过解释能够消解转移的说法便显得相当矛盾了，“主体从转移中的摆脱因而便被无限延期了”(E, 231).
这是否意味着拉康派的分析家们从来都不会解释转移？当然不是。拉康虽然断言“自然是要解释转移”(E, 271), 但他同时又对此种解释消解转移的力量不抱有任何的幻想。正如任何其他的解释一样，分析家必须运用他的全部技艺来决定是否以及何时要解释转移，而且尤其是必须避免让他的解释专门适合于解释转移。此外，他还必须知道自己通过这样的解释而试图实现的究竟是什么，不是旨在纠正病人与现实的关系，而是旨在维持分析的对话。“解释转移，这意味着什么呢？无非是用某种引诱来填补这一僵局的空隙。虽然它可能是欺骗性的，但是这个引诱会通过再度发动整个过程而服务于某种目的。”(Ec, 225)
在将转移描述为“正性”(positive)或“负性”(negative)的时候，拉康采取了两种取径。遵循弗洛伊德的观点，拉康有的时候会用这些形容词来指涉情感的本质，用“正性转移”来指涉那些爱的情感，而用“负性转移”来指涉那些侵凌性的情感 (Ec, 222)。然而，有的时候，拉康也会拿“正性”或“负性”这些术语来指涉转移对于治疗产生的那些有利的或不利的效果（见：E, 271: 拉康在此指出，当分析者的阻抗是反对暗示的时候，此种阻抗就必须被“放置在正性转移的等级上”，因为它维持了分析的方向）。
尽管拉康偶尔也确实会谈到反转移（COUNTERTRANSFERENCE), 但是他通常宁愿不去使用这个术语。
(transfert) The term 'transference'first emerged in Freud's work as simply another termfor the displacement of affect from one idea to another (see Freud, 1900a: SE V, 562). Later on, however, it came to refer to the patient's relationship to the analyst as itdevelops in the treatment. This soon became the central meaning of the term, and is thesense in which it is usually understood in psychoanalytic theory today.
The use of a special term to denote the patient's relationship to the analyst is justifiedby the peculiar character of this relationship. Freud was first struck by the intensity of thepatient's affective reactions to the doctor in Breuer's treatment of Anna O in 1882, whichhe argued was due to the patient transferring unconscious ideas onto the doctor (Freud, 1895d). As he developed the psychoanalytic method, Freud first regarded thetransference exclusively as a RESISTANCE which impedes the recall of repressedmemories, an obstacle to the treatment which must be 'destroyed' (Freud, 1905e: SE VII, 116). Gradually, however, he modified this view, coming to see the transference also as apositive factor which helps the treatment to progress. The positive value of transferencelies in the fact that it provides a way for the analysand's history to be confronted in theimmediacy of the present relationship with the analyst; in the way he relates to theanalyst, the analysand inevitably repeats earlier relationships with other figures (especially those with the parents). This paradoxical nature of transference, as both anobstacle to the treatment and that which drives the treatment forward, perhaps helps toexplain why there are so many different and opposing views of transferencenpsychoanalytic theory today.
Lacan's thinking about transference goes through several stages. His first work to dealwith the subject in any detail is 'An Intervention on the Transference' (Lacan, 1951), inwhich he describes the transference in dialectical terms borrowed from Hegel. Hecriticises ego-psychology for defining the transference in terms of AFFECTS;'Transference does not refer to any mysterious property of affect, and even when itreveals itself under the appearance of emotion, it only acquires meaning by virtue of thedialectical moment in which it is produced' (Ec, 225).
In other words, Lacan argues that although transference often manifests itself in theguise of particularly strong affects, such as LOVE and hate, it does not consist of suchemotions but in the structure of an intersubjective relationship. This structural definition of transference remains a constant theme throughout the rest of Lacan's work; heconsistently locates the essence of transference in the symbolic and not in the imaginary, although it clearly has powerful imaginary effects. Later on, Lacan will remark that iftransference often manifests itself under the appearance of love, it is first and foremostthe love of knowledge (savoir) that is concerned.
Lacan returns to the subject of the transference in the seminar of 1953-4. This time heconceives it not in terms borrowed from Hegelian dialectics but in terms borrowed fromthe anthropology of exchange (Mauss, Levi-Strauss). Transference is implicit in thespeech act, which involves an exchange of signs that transforms the speaker and listener:
In its essence, the efficacious transference which we're considering isquite simply the speech act. Each time a man speaks to another in anauthentic and full manner, there is, in the true sense, transference, symbolic transference-something which takes place which changes thenature of the two beings present.
(S1,109)
In the seminar of the following year, he continues to elaborate on the symbolic nature oftransference, which he identifies with the compulsion to repeat, the insistence of thesymbolic determinants of the subject (S2,210-11). This is to be distinguished from theimaginary aspect of transference, namely, the affective reactions of love and aggressivity In this distinction between the symbolic and imaginary aspects of transference, Lacanprovides a useful way of understanding the paradoxical function of the transference inpsychoanalytic treatment; in its symbolic aspect (REPETITION) it helps the treatmentprogress by revealing the signifiers of the subject's history, while in its imaginary aspect (love and hate) it acts as a resistance (see S4,135; S8,204).
Lacan's next approach to the subject of transference is in the eighth year of hisseminar (Lacan, 1960-1), entitled simply The Transference'. Here he uses Plato's Symposium to illustrate the relationship between the analysand and the analyst. Alcibiades compares Socrates to a plain box which encloses a precious object (Grkagalma); just as Alcibiades attributes a hidden treasure to Socrates, so the analysand seeshis object of desire in the analyst (see OBJET PETIT A).
In 1964, Lacan articulates the concept of transference with his concept of theSUBJECT SUPPOSED TO KNOW, which remains central to Lacan's view of thetransference from then on; indeed, it is this view of the transference which has come to beseen as Lacan's most complete attempt to theorise the matter. According to this view, transference is the attribution of knowledge to the Other, the supposition that the Other isa subject who knows;'As soon as the subject who is supposed to know existssomewhere... There is transference' (S11,232).
Although the existence of the transference is a necessary condition of psychoanalytictreatment, it is not sufficient in itself; it is also necessary that the analyst deal with thetransference in a unique way. It is this that differ-entiates psychoanalysis fromSUGGESTION; although both are based on the transference, psychoanalysis differs fromsuggestion because the analyst refuses to use the power given to him by the transference (seeE, 236).
From quite early on in the history of psychoanalysis it became common to distinguishbetween those aspects of the patient's relationship to the analyst which were adapted toreality'and those which were not. In the latter category fell all the patient's reactionswhich were caused by 'perceiving the analyst in a distorted way'. Some analysts used theterm 'transference'to refer to all aspects of the analysand's relationship to the analyst, inwhich case they distinguished the distorted 'neurotic transference'or 'transferenceneurosis'from the 'unobjectionable part of the transference'or 'therapeutic alliance' (Edward Bibring, Elizabeth Zeztel). Other analysts argued that the term 'transference'should be restricted to the 'unrealistic'or'irrational'reactions of the analysand (William Silverberg, Franz Alexander). However, the common assumption underlying both ofthese positions was that the analyst could tell when the patient was not reacting to him onthe basis of who he really was but rather on the basis of previous relationships with otherpeople. The analyst was credited with this ability because he was supposed to be better'adapted to reality'than the patient. Informed by his own correct perception of reality, theanalyst could offer 'transference interpretations'; that is, he could point out thediscrepancy between the real situation and the irrational way that the patient was reactingto it. It was argued that such transference interpretations helped the analysand to gain'insight'into his own neurotic transference and thereby resolve it or 'liquidate'it.
Some of Lacan's most incisive criticisms are directed at this way of representingpsychoanalytic treatment. These criticisms are based on the following arguments:
- The whole idea of adaptation to reality is based on a naive empiricist epistemology, involving an appeal to an unproblematic notion of 'reality'as an objective and self-evident given. This entirely neglects what psychoanalysis has discovered about theconstruction of reality by the ego on the basis of its own meconnaissance. Hence whenthe analyst assumes that he is better adapted to reality than the patient he has no otherrecourse than 'to fall back on his own ego'since this is the only 'bit of reality he knows' (E, 231). The healthy part of the patient's ego is then defined simply as 'the part thatthinks as we do' (E, 232). This reduces psychoanalytic treatment to a form of suggestionin which the analyst simply 'imposes his own idea of reality'on the analysand (E, 232). Thus 'the inability [of the analyst]to sustain a praxis in an authentic manner results, as isusually the case with mankind, in the exercise of power' (E, 226).
- The idea that the analysand's 'distorted perception of the analyst'could beliquidated by means of interpretations is a logical fallacy, since 'the transference isinterpreted on the basis of, and with the instrument of, the transference itself (S8,206). Inother words, there is no METALANGUAGE of the transference, no vantage pointoutside the transference from which the analyst could offer an interpretation, since anyinterpretation he offers 'will be received as coming from the person that the transferenceimputes him to be' (E, 231). Thus it is contradictory to claim that the transference can bedissolved by means of an interpretation when it is the transference itself which conditionsthe analysand's acceptance of that interpretation;'the emergence of the subject from thetransference is thus postponed ad infinitum' (E, 231).
Does this mean that Lacanian analysts never interpret the transference? Certainly not; Lacan affirms that 'it is natural to interpret the transference' (E, 271), but at the sametime he harbours no illusions about the power of such interpretations to dissolve thetransference. Like any other interpretation, the analyst must use all his art in deciding ifand when to interpret the transference, and above all mustavoidgearing his interpretations exclusively to interpreting the transference. He must also know exactlywhat he is seeking to achieve by such an interpretation; not to rectify the patient'srelationship to reality, but to maintain the analytic dialogue.'What does it mean, tointerpret the transference? Nothing else than to fill the void of this deadlock with a lure. But while it may be deceptive, this lure serves a purpose by setting off the whole processagain' (Ec, 225).
When describing the transference as 'positive'or 'negative', Lacan takes two differentapproaches. Following Freud, Lacan sometimes uses these adjectives to refer to thenature of the affects,'positive transference'referring to loving affects and negativetransference'referring to aggressive affects (Ec, 222). Sometimes, however, Lacan takesthe terms positive'and 'negative'to refer to the favourable or unfavourable effects ofthe transference on the treatment (see E, 271, where Lacan argues that when theanalysand's resistance opposes suggestion, this resistance must be placed in the ranks ofthe positive transference'on the grounds that it maintains the direction of the analysis).
Although Lacan does speak occasionallyof COUNTERTRANSFERENCE, hegenerally prefers not to use this term.