大卫·波德维尔—沃尔夫林和电影风格：关于电影诗学的一些想法
2021年逛大卫·波德维尔博客的时候，发现了他这篇关于电影风格/电影诗学的论文，当时闲来无事将此文翻译了出来。
前些日子大卫·波德维尔与世长辞，今将篇翻译发在豆瓣上，聊以纪念。
（侵删）
Wölfflin and Film Style: Some Thoughts on a Poetics of Pictures
沃尔夫林和电影风格：关于电影诗学的一些想法
大卫.波德维尔
无论是从训练还是从倾向上来看，我都觉得很难为海因里希·沃尔夫林在视觉艺术史上带来的许多哲学预设辩护。他寻求的是指导艺术史进程的规律。正是因为这些规律超越了艺术家个人做出的具体选择，所以他寻求推动这些规律的普遍原因。在这些原因中，民族性，或者说“种族”，是他在描绘德国和意大利艺术的差异时引用的一个概念。1
我并不相信他的这些说法，也不相信他的一些其他言论。然而，像许多艺术史学家一样，沃尔夫林留下了中级概念（middle‐level concepts）的遗产，这些概念可以帮助我们了解绘画、雕塑和建筑的重要特征。其中一些概念属于 "艺术理论 "的范畴：学校里教授的工艺技能，大师们流传下来的最佳范例，绘画附着阴影或设计三角形楣饰的 "规则"。通过研究这些规则，我们可以更好地把握特定时代和地点的艺术作品的设计目的。
其他概念是由研究者而非实践者发展起来的，旨在将艺术传统纳入更广泛的原则。亚里士多德的《诗学》就是包含这种倾向的最著名的例子。因此，让我们把“诗学”的方法称为将一个传统、流派或时期的特色艺术品的惯例系统化的努力。其中一些惯例将在艺术理论中得到阐述，而另一些则是完全默认的，并由研究人员揭示出来。
作为一名图像诗学家，沃尔夫林为视觉艺术的学生提供了大量的知识，尤其是对于长期以来对电影风格在历史上的连续性和变化充满好奇的电影研究者。在本文的第一部分，我将追溯这段历史的主流模式。一旦我们对电影风格发展达成了标准共识，我们就能更好地理解电影诗学是如何部分地借鉴沃尔夫林的概念，创造出对历史变化更丰富、更精炼的描述。
I. 无声电影的风格史
最早的电影史学家几乎完全忽视了德国语言艺术史中的视觉分析传统。甚至连欧文·潘诺夫斯基（Erwin Panofsky）和鲁道夫·爱因汉姆（Rudolf Arnheim）在撰写电影理论时，也没有发现应用希尔德布兰德（Hildebrand）、韦克霍夫（Wyckhoff）、瑞格尔（Riegl）和他们后继者的工作中的概念有什么成果。这是为什么呢？
首先，作家们无法通过接触电影来进行深入的研究。绘画和雕塑史学家们可以在闲暇时通过参观博物馆和其他资源库来研究原作。但在电影史最初的几十年里，研究者们没有档案可查。胶片拷贝几乎无法获得，用于观看胶片的机器也很罕见，而且价格昂贵。此外，虽然像沃尔夫林这样的艺术史学家可以出版原作的插图，但电影史学家只能接触到制作过程中拍摄的照片，而且这些照片通常不能复原屏幕上的影像。
20世纪20年代和30年代的电影史学家被迫放弃周密的分析，转而对连续性和变化进行概括。最有影响力的论述依赖于媒介的特性和目的论计划。如帕诺夫斯基所说："那么，在1905年之后，我们可以目睹一种新的艺术媒介逐渐意识到其合法性——即独特性——的可能性和局限性的迷人景象。 "2。
这里 "合法 "和 "独特 "的可能性是指电影区别于其他艺术形式的艺术资源。一部特定的电影不一定能利用该媒介所特有的美学资源。因此，这种艺术形式的美学史将以某种黑格尔式的方式，来展现媒介的可能性。
这些独特的可能性从本质上说是反戏剧的。电影的拥护者们认为，电影的发展使艺术逐渐从舞台上解放出来。电影在空间和时间上的自由使这种非语言媒介具有小说的范围和流动性。电影表演变得比舞台表演更自然，但也拥有不同风格化的方式：不是单纯的戏剧性默剧，而是更巧妙、更有机的东西。特写，这种在舞台上不可行的技术，不仅加强了表演，而且使放大的、移动的人脸成为新的审美对象。特写镜头还使整个可见的世界，从一朵花到转动的门把手，都成为戏剧的一部分。
剪辑技术使这些创造性的选择成为可能。通过从远景切换到近景，电影制作者可以放大一张脸、一双手或桌上的一把刀。通过从一处剪切到另一处，电影制作者可以扩大戏剧的舞台，加强悬念，就像大卫·格里菲斯（D. W. Griffith）所做的那样，在处于危险中的无辜者和匆忙赶到现场的救援人员之间进行交叉剪辑。通过插入呈现梦境或记忆的画面，或者简单地切换成主观镜头，电影制作者就可以暴露人物的内心活动。剪辑也使时间具有可塑性。制作者可以跳过几个小时或几年，甚至像格里菲斯在《党同伐异》（Intolerance，1916）中做的那样，交叉切换不同的时代来创造历史的相似性。早在1915年，雨果·明斯特伯格（Hugo Münsterberg）就指出，电影的这些能力使它像人类的注意力、记忆力、反思力等心理过程一样，具有流动性和广泛性。
到默片时代结束时，以剪辑为基础的技术已经成为商业故事片和大量纪录片的通用语言。用沃尔夫林的话说，一种广泛的群体风格出现了，它不与特定的国家甚至个人挂钩。潘诺夫斯基所说的媒介对其资源的 "醒悟"，并不是简单粗略的表达方式。在很大程度上，早期的电影史学家认为，在无声时代，特定的艺术家之所以变得重要，主要是因为他们在相当明确的阶段展露了电影独特的艺术资源，并让电影脱离了戏剧的模式。
还有乔治·梅里爱（Georges Méliès）的 "人工置景 。"3 尽管很接近舞台剧，但这些单镜头的场景通过剪辑连接起来可以就讲述一个故事，就像《月球旅行记》（A Trip to the Moon，1902）那样。还有埃德温·鲍特（Edwin S. Porter），他在《火车大劫案》（The Great Train Robbery，1903）中利用剪辑来表现一个拥有多条线索的故事。尤其是格里菲斯，据说他在1908年至1916年间发现了电影剪辑最基本的资源。《一个国家的诞生》（The Birth of a Nation，1915）据称是格里菲斯的集大成之作，为20世纪20年代的苏联蒙太奇电影人奠定了基础，他将剪辑的节奏感、戏剧性和概念化的可能性发展到了顶点，爱森斯坦的《战舰波将金号》（Battleship Potemkin，1925）通常被认为是最好的例子。
在采用媒介发展的模式时，电影史学家与学术性艺术史并不完全同步。当然，艺术史学家长期以来赋予某些大师以创新者的地位，比如对线性透视的描述。但许多历史学家倾向于将艺术家成就的质量相对化。例如，沃尔夫林就不厌其烦地指出，巴洛克不应该被认为是向更伟大的模仿力或绝对的美学观念迈进的一步。没有一个时代的风格天生就优于另一个时代。乔托没有画阴影，但这并不等同于粗糙简陋，因为他不需要用阴影来实现他所追求的东西。4相比之下，20世纪20年代的电影剧作家对20世纪初的 "戏剧性 "电影只是嗤之以鼻。在他们眼中，它并不是一种有自身审美目的的电影制作模式，而是一个原始阶段，如果电影要成为一门艺术，就必须超越它。
二、有声电影合成学与新风格的兴起
战后法国评论家安德烈·巴赞对这个标准的叙述进行了细微的调整。他认为声音的出现给许多无声电影的技巧赋予了价值。在几乎没有明显干预的情况下，有声电影可以清晰地呈现出故事的情境。这种电影仍然用到了剪辑技术，但比格里菲斯的交叉剪辑或爱森斯坦的理性蒙太奇谨慎得多。特写镜头被最小化，镜头被允许长时间持续进行而不必担心 "戏剧性"。一种融合了不具有冲突性的剪辑、摄影机运动和透明的舞台风格在20世纪30年代成为世界电影制作的主流。正如巴赞所说，全世界的电影人都决定尊重现实的时空连续性。
实际上，巴赞提出，1939之前的电影史可以分为三个时期的风格：对戏剧表演的简单记录、1910年代末到1920年代的双线剪辑、以及国际有声电影的平衡综合。但是，最后一种风格还没有成为常态的时候就受到了其他风格的挑战，这种风格构成了另一个阶段的进步，即 "电影语言的辩证发展。"6巴赞像他的前辈一样把这种启示追溯到创新性电影人的作品中，如让·雷诺阿（Jean Renoir）、奥逊·威尔斯（Orson Welles）和威廉·惠勒（William Wyler）。
这些电影制作者将剪辑减少到了一个不寻常的程度，整个场景可能只通过一个镜头演绎。摄影机的移动起到了剪辑的作用，移动镜头或推拉镜头可以带领观众穿越空间、进入情境。即使没有摄影机的移动，静止的画面也能在左右景、前后景之间产生戏剧性张力。这种新风格的核心，便是巴赞所谓的景深（profondeur de champ）， 也通常被翻译为“深焦”。 《游戏规则》（The Rules of the Game ，1939）、《公民凯恩》（Citizen Kane ，1941）、《小狐狸》（The Little Foxes，1941）等影片表明，通过延长镜头长度，可以形成纵深的舞台效果，与 "固有的电影 "剪辑技术所产生的效果一样尖锐有力。这些电影实际上恢复了 "原始 "技术，但却赋予了它前所未有的意义。
对巴赞来说，这不仅仅是技术上的发展。就像早期的史学家称赞格里菲斯、俄国人及其他揭示了电影真实本质的人一样，现在巴赞也为他的代表性导演们做了同样的事情。他们从电影真正的本质出发，直观地创造艺术，在不间断的时间里记录运动。例如，雷诺阿 “超越了蒙太奇所提供的资源，因此发现了一种电影形式的秘密，这种电影形式可以让人们畅谈一切，而不必把世界切分成小碎片”7。
这个标准的叙述除了注意到诸如交叉剪辑或特写镜头等特殊技术外，几乎没有提供风格批评的依据。8 通过引入另一种时代风格，即1940-1950年代的深焦长镜头，巴赞鼓励了沃尔夫林在其艺术史作品中发展的比较研究法。批评家们可以将剪辑主导的处理方式与舞台主导的处理方式进行对比，有时是通过实际的例子，有时是通过想象如何以不同的方式呈现同一场景。巴赞就在《伟大的安巴逊》(The Magnificent Ambersons，1942)的厨房场景中做了一个著名的单镜头尝试，他指出在高潮时刻切入范妮阿姨的特写会破坏威尔逊的目的。延长的长镜头将我们的注意力集中在她的侄子乔治狼吞虎咽的食物上，所以她的泪水爆发得异常有力（图1）。
然而，巴赞面临着与默片时代结束时以剪辑为基础的叙事的相同问题。一旦导演们发现了电影的真正使命，就不会出现新的群体风格。一些导演将坚守于有声电影的综合，一些导演会跟随威尔斯、雷诺阿和惠勒的脚步。也可能会出现个人风格，例如与马克斯·奥菲尔斯（Max Ophuls）对推轨镜头的依赖或希区柯克复原无声电影的装置来制造悬念，但大规模的风格变化似乎已经结束。巴赞预见到的电影变化主要发生在类型、技术（彩色、宽银幕）和题材领域。剩下的是对导演批评的鉴赏力，即在既有的传统中找出更细微的差异。
三、沃尔夫林与修正主义电影史
即使在今天，大多数关于电影史的流行描述都包含了标准叙述和修正的巴赞派的一些变体。但在更广泛的文化中还没有被完全记录下来的，是20世纪70年代的研究者们开始深刻反思的电影风格史。档案馆向学者敞开了大门，大学开设了电影研究的高级课程，出版商乐意出版研究电影艺术传统的书籍。
这其中最重要的影响是拓宽了电影典籍的范围。当学者们能够看到官方经典之外的电影样本时，他们开始意识到，巴赞重塑但未否定的标准叙述简化了记载。最关键的是，研究者们必须面对1900年代和1910年代电影制作惊人的多样性。在20世纪70年代开始学者们的努力下，早期电影研究成为电影研究中最成熟的领域之一。
尽管如此，电影史学家却出奇地不愿利用也许能够阐明风格传统的艺术史模型来分析模式和原理。然而，在我看来，立足于诗学的研究方法似乎卓有成效，沃尔夫林提供的重要指导，使我们能够把握电影制作者的创造性选择。为了说明这一点，我想表明，《艺术史原理》中提出的平面构图和纵深构图之间的巨大对比，可以使我们认识到早期历史学家所忽略或误解的电影风格。
不过，我们首先要注意一些问题。沃尔夫林明确表示，他著名的二元性是有些理想化的，而且人们总是可以发现二者混合和介于中间的例子。10他意识到，自己通过两两比较的方式进行的阐述，源于使用双幻灯片的教学方法，这种方法可能会夸大图像之间的对比。11而此方法在电影中的应用就更复杂了，因为一部电影可能包括混合构图原则的镜头。如果我们要利用平面/纵深的对比，我们应该认识到最初的探索是相当粗略的，需要进一步的研究以获得精确度。
四、作为艺术史的电影史：舞台造型传统
沃尔夫林称，平面构图是16世纪意大利艺术和德国艺术的特点。它们将主要的人物和动作排列成一排，视点通常与后方平面成直角。就算有深度，较远的层次也会以带状或条状的形式呈现。但这种视觉阵列并非简单地排列起来并面对我们的图形“原始”构造。经典的平面构图已经掌握了前短后长的技巧，因此，虽然重要的人物往往被设置在同一平面上，但身体和脸部的角度却可以是多种多样的。其原型是1495-1498年列奥纳多（Leonardo）的《最后的晚餐》。
相比之下，纵深式构图成为了16世纪巴洛克艺术的特点。在这种构图里，重要的动作被延展到深度的位置，通常可以形成强烈的大规模对比。1614年鲁本斯（Rubens）的《哀歌》就是一个生动的例子，在这幅画中，基督死去的身躯从画面上陡然向后倾斜，送葬者在他周围呈对角线排列。
虽然巴赞没有参考《艺术史原理》，但他对威尔斯和惠勒的描述却很符合纵深风格的概念12。20世纪40年代的许多美国电影在整个过程里都在构建对角线构图，人物沿着锐利的焦点平面排列。最生动的地方是沃尔夫林所说的 "夸张的前景"，在巴洛克艺术中，它 "增加了透视缩小的突然性。"13(见图2.)沃尔夫林强调了其他有利于设置深度的特征，例如走廊、黑暗的前景和画面内的画框。这些技术在1940年代的好莱坞电影里也能找到，令我们倾向于给这类电影贴上“巴洛克”的标签。
不过，我们还可以再往前追溯。沃尔夫林的 "纵深"概念提醒我们注意另一种直到最近才被电影史学家所察觉的趋势。无声电影史学家曾谴责大多数格里菲斯之前的电影在本质上是戏剧性的，在很大程度上，修正的巴赞派也持同样的观点。14 但我们可以从整个1906-1920年期间的欧洲和一些美国电影制作中发现复杂的绘画策略。这是一个 "舞台造型"电影制作的时代，这个词不仅味着类似绘画的设计，也意味着与当时的舞台 "电影 "的亲近关系。15
早期的电影舞台可以表现出纵深性和平面性两种特质。一部电影通常呈现出景深较浅的内景和倾斜度很大的外景（图3和图4）。当电影播放到10分钟或往后时，导演们就会延长他们的场景，并发展出更精致的室内布景和更复杂的舞台布置。他们很少在场景内进行剪接，但在延长的镜头中，导演们探索出了大量的构图方法，以引导观众的注意力。
有些时候，舞台造型明显是平面的，人物呈平行状排列。导演可能会通过移动镜头或隐藏的策略来突出一个动作，让观众从侧面的缝隙中瞥见一个重要的面孔或物体（图5）。但舞台造型也承载了许多对纵深镜头设计的探索。《刺杀吉斯公爵》（1908）便是早期一个著名的例子。它被宣传为具有壮观的剧场高度，但其构图显示了前景和后景之间的动态作用。其中一个相当拥挤的场景大量使用了沃尔夫林的 "巴洛克式的布景和装饰"。（图6）
在1910年代的纵深风格中，有时画面很宽广，深度以中间占有大片区域的人物为标志（图7）。其他时候画面很紧密，人物在情节的进行中展现和填补空白（图8和图9）。画面通过各种线索引导我们注意重要的线速：灯光、人物的中心、正面描绘的人物、向前景的推进和移动的速度。离镜头较近的人物可以遮住或显示较远的人物，这种编排在1910年代已经变得相当复杂（图10和图11）
乍一看，这些对画面空间的处理似乎与人们在剧院舞台上看到的并无不同。但这一时期的电影人很清楚，电影的表现空间并不像舞台表演那样。
传统的舞台装置使戏剧表演空间既宽泛又浅显。水平表演的动作使观众席上的观众都能领会演员的行动。相比之下，电影的表现空间是由镜头的光学特性创造的。虽然我们像是在一个剧院的盒子里观看 ，但实际上摄影机捕捉到的区域是一个金字塔，呈直角倾斜，在镜头表面逐渐变小。这种表现的区域很窄但很深。因此，电影里的动作必须以唯一重要的视线—也就是摄影机的视线来进行编排。这里展示的许多镜头都无法在舞台上表现，因为观众席的某些座位看不到主要的活动区域。
电影独特的表现空间，让舞台造型风格丰富的空间操作得以实现。具有讽刺意味的是，这种对移动身体的透视捕捉，就像剪辑一样被认为是 "电影所特有的 "。
此外，舞台造型的方法远非纯粹的戏剧记录，而是构成了一种独特的群体风格。各个电影人虽然以不同的方式进行了探索，但他们都在用相同而广泛的舞台模式进行工作。许多电影人以相当纯粹的形式来表现这种风格，单镜头的场景只用字幕来切断。其他电影人则通过剪辑选择调整画面——通常是通过轴向切入来扩大画面空间的一部分。这种 "场景插入 "的方式是1910年代后期电影风格的特点。有时，我们也会遇到一些通过运用非常近的前景来将画面发挥到极致的电影人。罗伯特·莱纳特（Robert Reinert）的《神经质》（Nerven，1920）中的一个场景就类似于我们在威尔斯和惠勒的电影中看到的大脑袋（图12）。17
我们可以将舞台造型看作是一种时代的风格，因为在1920年前后，与之同时发展起来的以剪辑为主导的风格几乎完全取代了舞台造型，风靡了全世界的电影院。然而，舞台造型时代被突出的纵深手法，在20世纪20年代和30年代的一些电影，尤其是俄罗斯电影的镜头设计中被放大了。而这一时期的舞美设置也成为电影制作永久的资源。自20世纪60年代以来，它们在长镜头电影中显得尤为突出。
五.、电影史与艺术史：平面风格
电影作品也展现了一段平面构图的历史。我们发现，在故事片制作的头十年左右，平面构图就以明显的形式出现了（图13），正如我们所见到过的，它在戏剧舞台时代的形式更为复杂。在经典的默片和20世纪30年代的电影中，大多数戏剧性的相遇都采用了一种较为温和的平面构图。画面中的人物在同一平面上以侧面或头部稍作旋转产生四分之三的视角面对彼此（图14）。镜头/反打镜头的剪辑将这种对峙转化为一系列的镜头。现在，每一个演员都以四分之三的视角呈现，而回应的镜头则提供了另一个演员的呼应影像（图15和16）。但两者通常会被认为存在于同一平面上，特别是当背景表面与摄像机垂直时。
从20世纪40、50年代开始流行的巴洛克式纵深构图，是国际电影温和的平面布局的主要替代方式，虽然威尔斯、惠勒等人的大前景难以在彩色和宽银幕电影中出现，但一些电影人还是设法使其得到了实现（图17）。让人意想不到的是，20世纪60年代开始出现了更为极端的平面化造型。
导演们开始将摄影机垂直设置在背景的表面。人物站成一排，面向观众或转过九十度（图18和19）。整个阵列被平铺放置以弱化深度。就像许多国家出现的其他风格趋势一样，这种仿造的平面造型在随后的几十年里成为国际电影的一种选择。它可以用于群像，也可以用于个人（图20和21）。20世纪90年代，北野武将其作为电影的核心（图22），到2014年，韦斯·安德森则将它的变体确立为自己的标志性风格（图23和24）。
沃尔夫林由于掌握了前缩透视法的技术革新，十六世纪意大利艺术复杂的平面风格不再同于“原始”的变体。同样，电影的新平面风格也只有在剪辑允许我所说的罗盘点切割之后才能成功：180度的反转来显示出人物所看的东西（通常是面对镜头的其他人物）。
六、阐释
此外，沃尔夫林的参考框架有利于我们发现和描述电影表现手法的一些重要变化。但如何解释这些变化呢？我们不需要移用一些沃尔夫林认为合适的解释，譬如对民族性的吸引力。我们也不需要接受他的另一个主要的因果机制，即 "视觉本身是有历史的。"18这一点并不像表面上看起来那么突出：他指出，比 "视觉模式 "更好的术语应该是 "想象模式 "19，然而鉴于在短短的几十年里有那么多的电影风格共存，这个划时代的概念并非电影研究的合理选择。
另一位图像诗学家贡布里希（E. H. Gombrich）提出了更合理的阐释框架。沃尔夫林倾向于方法论的整体主义，而贡布里希则认为形式的发展主要可以通过方法论的个人主义来解释。"艺术家总是面临着这样的问题：'有什么是我可以做的？’"20此外，贡布里希赋予艺术图式的概念以比沃尔夫林有更具体的存在。它们较少存在于艺术家的头脑中，而更多地存在于模板书和初学者在训练中遇到的传统图示解决方案中。最有力的地方是，贡布里希拒绝将民族性和 "种族 "作为互为因果的关系，而是将重点放在时尚、竞争和艺术在特定艺术群体中的社会功能上。
我发现，当研究一门依赖于大众市场的艺术，一个紧密联系的创作者群体以及声音、图像和叙事的既定规范时，从问题和解决方案以及模式和修正方面构思的框架更具有启示性。据此，我认为我们应该寻找贡布里希式序列的原因：生产条件、影响力、时尚、竞争和区分自己作品的冲动。例如，复杂的舞台造型风格可能起源于戏剧和绘画模式，但它们经过反复试验才转变成符合摄影机光学透视的东西。电影制作者可以从透视绘画中借用图形模式，并根据运动和摄影机位置的限制来塑造它们。
在后来的几年里，我认为有一连串的因素造成了极端的平面影像的兴起。20世纪60年代，一种对外景拍摄的新依赖促使电影人们使用长镜头，从而产生了景深较浅的影像。画面的 "扁平化"以及新的宽银幕比例，鼓励了一些电影人，特别是那些倾向于拍摄长镜头的电影人从正面和侧面来设置动作。这也有一种表现上的优势。这种模式可以进一步促进艺术电影的 "去戏剧化 "特征，它为好莱坞浮躁的活力提供了一个鲜明的替代方案。
但是，我们应该记住沃尔夫林的建议，即这些模式没有固有的表达特性，它们在本质上主要是装饰性的。舞台造型风格既适合喜剧又适合情节剧，而在美国主流娱乐类型中，仿造的平面外观既适合滑稽剧，也适合正剧。 总而言之，沃尔夫林给我们留下了中级概念的遗产，这些中级概念有助于详细说明和分析电影制作中的绘画传统。然而，他的作品也提醒我们视觉现象的形式有多么复杂，以及这种形式是如何难以得到阐释的。
参考文献
1.The purported contrasts between German and Italian national character are developed in detail in Heinrich Wölfflin, The Sense of Form in Art: A Comparative Psychological Study, trans. Alice Muehsam and Norma A. Shetan (New York: Chelsea House, 1958). A typical remark is that the painterly tendency (as opposed to the linear one) “runs in the veins of the Germanic race.” Principles of Art History: The Problem of the Development of Style in Later Art, 7th edition [1923], trans. M. D. Hottinger (New York: Dover, 1950), p. 67.
2.Erwin Panofsky, “Style and Medium in the Motion Pictures,” in Three Essays on Style, ed. Irving Lavin (MIT Press, 1995), pp. 91–125, at p. 108.
3.See John Frazer, Artificially Arranged Scenes: The Films of George Méliès (Boston: G. K. Hall, 1979).
4.Heinrich Wolfflin, Classic Art: An Introduction to the Italian Renaissance, trans. Peter and Linda Murray (London: Phaidon, 1952), p. 5.
5.The argument is set out most fully in André Bazin, “The Evolution of the Language of Cinema,” in What Is Cinema? vol. I, ed. and trans. Hugh Gray (University of California Press, 1967), pp. 23–40.
6.Bazin, “Evolution of the Language of Cinema,” p. 35.
7.Bazin, “Evolution of the Language of Cinema,” p. 38.
8.Examples are Lewis Jacobs's analyses of crosscutting in sequences of Griffith's Birth of a Nation and Intolerance in The Rise of the American Film: A Critical History (Columbia University Teachers College Press, 1968), pp. 180–198.
9.The relevant argument is found in Wölfflin, Principles of Art History, chap. II.
10.See Wölfflin, Principles of Art History, pp. viii, 14, 233; and Heinrich Wölfflin, “‘Kunstgeschichtliche Grundbegriffe’: eine Revision,” in Gedanken zur Kunstgeshichte (Basel: Schwabe, 1941), pp. 18–24, at p. 23.
11.Wölfflin, The Sense of Form in Art, p. 4.
12.Principles appeared in French translation in 1952, some years after Bazin's initial discussions of deep focus. It is likely that Bazin was sensitized to the new style by 1940s press material signed by Wyler, Welles, and cinematographer Gregg Toland. See David Bordwell, On the History of Film Style (Harvard University Press, 1997), pp. 56–57.
13.Wölfflin, Principles of Art History, p. 84.
14.True, Bazin claimed to find a “primitive” profondeur de champ in early cinema, but his example, an undated image from the Onésime series, is striking exactly because of its similarity to the aggressive foregrounds of Welles. See Bordwell, On the History of Film Style, pp. 62–63.
15.Major investigations of tableau cinema are Ben Brewster, “Deep Staging in French Films 1900–1914,” in Early Cinema: Space Frame Narrative, ed. Thomas Elsaesser (London: British Film Institute, 1990), pp. 45–55; Yuri Tsivian, “Portraits, Mirrors, Death: Death: On Some Decadent Clichés in Early Russian Films,” Iris 14–15 (1992): 70–78; Tom Gunning, “Notes and Queries about the Year 1913 and Film Style: National Styles and Deep Staging,” 1895, special issue: “L'année l9l3 en France” (1993): 195–204; Ben Brewster and Lea Jacobs, Theatre to Cinema (Oxford University Press, 1997), pp. 164–187. The account of the tableau style and planimetric staging that follows is drawn from Bordwell, On the History of Film Style, pp. 158–207, and David Bordwell, Figures Traced in Light: On Cinematic Staging (University of California Press, 2005). See also entries on my website, www.davidbordwell.net.
16.Wölfflin, Principles of Art History, p. 86.
17.For more on Reinert, see my essay “Taking Things to Extremes: Hallucinations Courtesy of Robert Reinert,” in Poetics of Cinema (New York: Routledge, 2008), pp. 281–326.
18.Wölfflin, Principles of Art History, p. 11.
19.Foreword to the sixth edition, Principles of Art History, p. vii. It seems clear that “modes of vision” are not purely optical phenomena. Wölfflin explains that the artist internalizes “what choice of formal possibilities the epoch had at its disposal.” This fits the famous claim that immediately precedes the statement about vision's history: “Every artist finds certain visual possibilities before him, to which he is bound. Not everything is possible at all times” (Principles of Art History, p. 11). In the 1933 essay “‘Kunstgeschichtliche Grundbegriffe’: eine Revision,” Wölfflin explains that he used the term “mode of vision” in a sense analogous to speaking of “the artist's ‘eye’ and his ‘vision,’ by which we mean the manner in which objects take form in an internal representation” (Gedanken zur Kunstgeshichte, pp. 19–20).
20.Quoted in Ernst Gombrich and Didier Eribon, Looking for Answers: Conversations on Art and Science (New York: Abrams, 1993), p. 168.
21.This duality is not quite as sharp as I am making it out. Wölfflin, for instance, suggests that one major source of the artist's “mode of beholding” is prior art, “the effect of picture upon picture”—a tenet of Gombrich's thinking as well. See Wölfflin, Principles of Art History, pp. 229–230; E. H. Gombrich, Art and Illusion: A Study in the Psychology of Pictorial Representation (Princeton University Press, 1969), pp. 63–90.
原文 By both training and inclination, I find it hard to defend many of the philosophical presuppositions Heinrich Wölfflin brings to bear on the history of the visual arts. He seeks laws governing the course of art history. Because these laws transcend the concrete choices made by individual artists, he seeks large‐scale causes propelling them. Among these causes is national character, or “race,” a concept he invokes in charting differences between German and Italian art.1
These and other of his claims seem to me implausible. Yet like many art historians, Wölfflin leaves a legacy of middle‐level concepts that can help us notice important features of painting, sculpture, and architecture. Some of these concepts fall under the rubric of “art theory”: the craft skills taught in schools, the best practices passed down from masters, the “rules” for painting attached shadows or designing pediments. By studying these rules we can better grasp what artworks in particular times and places are designed to do.
Other concepts, developed by researchers rather than practitioners, aim to subsume artistic traditions to broader principles. Aristotle's Poetics stands as the most famous example of this tendency. So let us call the “poetics” approach an effort to systematize the conventions of a tradition, a genre, or a period's characteristic artworks. Some of these conventions will be articulated in art theory, while others are wholly tacit and are brought to light by the researcher.
In the role of a pictorial poetician, Wölfflin offers a great deal to the student of the visual arts. In particular, we film researchers have long been curious about continuity and change in cinematic style across history. In the first section of this essay, I trace the dominant model of that history. Once we have a sense of the standard consensus about cinema's stylistic development, we are in a better position to see how a cinematic poetics, drawing in part on Wölfflin's concepts, can create a richer and more refined account of historical change.
I. STYLISTIC HISTORY OF SILENT CINEMA
The earliest historians of cinema almost completely ignored the tradition of visual analysis in German‐language art history. Even Erwin Panofsky and Rudolf Arnheim, when they wrote about film, did not find it fruitful to apply concepts from the work of Hildebrand, Wyckhoff, Riegl, and their successors. Why?
For one thing, writers had no access to films for close scrutiny. Historians of painting and sculpture could study the original works at leisure by visiting museums and other repositories. But during the first decades of film history, there were no archives accessible to researchers. Film copies were virtually unobtainable, and machines for viewing them were rare and costly. Moreover, while art historians like Wölfflin could publish illustrations of original works, film historians had access only to photographs made during production, which usually did not replicate the images on‐screen.
Film historians of the 1920s and 1930s were forced to sacrifice close analysis to generalizations about continuity and change. The most influential account relied upon medium specificity and a teleological plan. Panofsky put it well: “After 1905, then, we can witness the fascinating spectacle of a new artistic medium gradually becoming conscious of its legitimate—that is, exclusive—possibilities and limitations.”2
“Legitimate” and “exclusive” possibilities here denote artistic resources that distinguish cinema from other art forms. A particular movie does not necessarily exploit the aesthetic resources specific to the medium. Accordingly, the aesthetic history of that art form will be, in a somewhat Hegelian fashion, the unfolding manifestation of the medium's possibilities.
Those distinct possibilities were, at bottom, anti‐theatrical. Cinema's development, its advocates claimed, gradually emancipated the art from that of the stage. Cinema's freedom in space and time gave this nonverbal medium the range and fluidity of the novel. Cinematic acting became more naturalistic than stage performance, but also stylized in a different way: not mere theatrical pantomime but something subtler and more organic. The close‐up, a technique not feasible on the stage, not only enhanced the performance but made the enlarged and mobile human face a new object of aesthetic apprehension. Close‐ups also allowed the entire visible world, from a flower to a turning doorknob, to become part of the drama.
What made most of these creative options possible was the technique of editing. By cutting from a distant shot to a nearer view, the filmmaker could enlarge a face, a pair of hands, or a knife on the table. By cutting from place to place, the filmmaker could expand the arena of the drama and intensify suspense, as D. W. Griffith did in crosscutting between innocents in danger and the rescuers hastening to the scene. By inserting images presenting dream or memory or simply by cutting to optical point‐of‐view shots, the filmmaker could expose the inner life of the characters. Cutting also made time malleable. The filmmaker could skip over hours or years or even, as Griffith did in Intolerance (1916), intercut different epochs to create historical parallels. As early as 1915, Hugo Münsterberg pointed out that these capacities of cinema made it as fluid and wide‐ranging as the human mental processes of attention, memory, and reflective thought.
By the end of the silent era, editing‐based technique was the lingua franca of commercial fictional cinema and a good deal of documentary as well. In Wölfflin's terms, there emerged a broad group style that was not tied to particular nations or even individuals. Panofsky's reference to the medium's “becoming conscious” of its resources was not simply shorthand. To an important extent, early historians of cinema held that in the silent era specific artists became important chiefly because they disclosed, in fairly clear‐cut stages, film's unique artistic resources and carried it away from theatrical modes.
There was Georges Méliès with his “artificially arranged scenes.”3 Despite being stage‐like tableaus, these single‐shot scenes could tell a story when they were linked by cuts, as in A Trip to the Moon (1902). There was Edwin S. Porter, who in The Great Train Robbery (1903) used editing to present a story with several strands of action. Above all there was Griffith, who was said to have discovered the basic resources of film editing between 1908 and 1916. Griffith's achievement, purportedly crystallized in The Birth of a Nation (1915), was foundational for the Soviet montage filmmakers of the 1920s, who developed the rhythmic, dramatic, and conceptual possibilities of editing to their highest pitch. Eisenstein's Battleship Potemkin (1925) was usually considered the prime example.
In adopting the model of the medium's progress, film historians were not exactly in synchronization with academic art history. Of course art historians had long granted certain masters the status of innovators, as in accounts of linear perspective. But many historians were inclined to relativize the quality of artists’ accomplishments. Wölfflin, for instance, was at pains to indicate that the baroque should not be considered a step toward greater imitative power or an absolute idea of beauty. No one epoch's style was inherently superior to another. Giotto did not paint shadows, but this was not crudity; he did not need them to achieve what he sought.4 By contrast, the film writers of the 1920s had only scorn for the “theatrical” cinema of the early 1900s. It was not a mode of filmmaking with its own aesthetic purposes but rather a primitive phase that had to be surpassed if cinema was to become an art in its own right.
II. THE SOUND FILM SYNTHESIS AND THE RISE OF A NEW STYLE
The standard story was nuanced by the postwar French critic André Bazin. He argued that the arrival of sound had made many valorized techniques of the silent film dépassé. Sound films presented a story's situations lucidly and with little overt intervention. Editing persisted, but it was more discreet than Griffith's crosscutting or Eisenstein's intellectual montage. Close‐ups were minimized, and shots were allowed to run long with no fear of “theatricality.” A style blending unaggressive editing, camera movement, and transparent staging came to dominate world filmmaking during the 1930s. As Bazin put it, filmmakers around the world had decided to respect the spatiotemporal continuity of reality.5
Bazin in effect proposed that up to 1939 film history had three period styles: the primitive recording of theatrical performances, the hyperbolic editing of the late 1910s through the 1920s, and the balanced synthesis of international sound cinema. But no sooner had this last style become normal than it was challenged by a style that constituted another phase of progress, a “dialectical step forward in film language.”6 Like his predecessors, Bazin traced this revelation to the work of innovative filmmakers: Jean Renoir, Orson Welles, and William Wyler.
These filmmakers minimized editing to an unusual degree. Entire scenes might be played in one shot. Camera movement did duty for cutting, with panning or tracking taking us across spaces and into scenes. Even without camera movement, the static frame could create dramatic tension between right and left, foreground and background. What Bazin called profondeur de champ, usually translated as “deep focus,” was central to the new style. The Rules of the Game (1939), Citizen Kane (1941), and The Little Foxes (1941) showed that staging in depth, supplemented by an extended shot duration, could be shaped to provide effects as pointed and powerful as anything yielded by the “inherently cinematic” technique of editing. These films were, in effect, returning to “primitive” technique but investing it with a significance it had not had.
For Bazin, this was no mere technical development. Just as earlier historians had praised Griffith, the Russians, and the rest for disclosing the true nature of cinema, now Bazin did the same for his emblematic directors. They were intuitively making art from cinema's real essence, the recording of movement in uninterrupted time. Renoir, for instance, looked “back beyond the resources provided by montage and so uncovered the secret of a film form that would permit everything to be said without chopping the world up into little fragments.”7
The standard story had provided little basis for stylistic criticism, apart from noting privileged techniques like crosscutting or close‐ups.8 By introducing an alternative period style, that of 1940s–1950s deep‐focus long takes, Bazin encouraged the sort of comparative method that Wölfflin had developed in his art‐historical work. Critics could contrast the editing‐driven handling with staging‐driven ones, sometimes by using actual examples, sometimes by imagining how the same scene could be presented in alternative ways. Bazin did this famously with the single‐shot kitchen scene in The Magnificent Ambersons (1942), pointing out that a cut‐in close‐up of Aunt Fanny at the climactic moment would have damaged Welles's purpose. The extended long shot concentrates our attention on her nephew George wolfing down his food, so her tearful outburst comes with exceptional force (Figure 1).
Bazin was confronted, however, with the same problem that arose with editing‐based accounts at the close of the silent era. Once directors had discovered cinema's true vocation, there would be no new group style. Some directors would persist in the sound‐film synthesis, and some would follow the lead of Welles, Renoir, and Wyler. Individual styles could emerge, as with Max Ophuls's reliance on tracking shots or Hitchcock's reversion to silent‐film devices for the sake of suspense, but large‐scale stylistic change seemed to have ended. Mostly, Bazin foresaw cinematic change taking place in the realm of genre, technology (color, widescreen), and subject matter. What remained was the connoisseurship of auteur criticism, identifying finer‐grained differences within established traditions.
III. WÖLFFLIN AND REVISIONIST FILM HISTORY
Even today, most popular accounts of film history embrace some variant of the standard story and the Bazinian revision. What has not fully registered in the wider culture is that in the 1970s, researchers began profoundly rethinking cinema's stylistic history. Archives opened their doors to scholars, universities launched advanced programs in film research, and publishers began welcoming books treating film's artistic traditions.
Of greatest importance was the widening of the canon. When scholars could see a sampling of films beyond the official classics, they began to realize that the standard story, which Bazin recast but did not repudiate, simplified the record. Most crucially, researchers had to face the astonishing variety of filmmaking in the 1900s and 1910s. Thanks to the efforts of scholars who began work in the 1970s, early cinema studies became one of the most mature domains of cinema studies.
Nonetheless, film historians have been surprisingly reluctant to draw upon art‐historical models for analytical models and principles that might illuminate stylistic traditions. Yet an approach grounded in poetics seems to me fruitful, and Wölfflin offers some important pointers in enabling us to grasp the creative options available to the filmmaker. To illustrate, I want to show that the formidable contrast launched in Principles of Art History between planimetric composition and recessional composition can lead us to recognize aspects of film style that were missed, or misconstrued, by earlier historians.9
First, though, some caution. Wölfflin was explicit that his famous dualities were somewhat idealized, and that one could always find mixtures and intermediate examples.10 He was aware that his exposition through pairwise comparison, born of a pedagogy using dual lantern slides, could exaggerate contrasts between the images.11 Application to cinema is even more complicated, since a film may include shots mixing compositional principles. If we are to make use of the plane/recession polarity, we should recognize that our initial soundings would be fairly rough and need further research to gain precision.
IV. FILM HISTORY AS ART HISTORY: THE TABLEAU TRADITION
Planimetric composition, Wölfflin claims, is characteristic of Italian and German art of the 1500s. Here the major figures and action are arranged in a row, with the viewpoint typically at right angles to the rear plane. If there is depth, the more distant layers are also presented as bands or strips. But the visual array is not a “primitive” construction of figures simply lined up and facing us. Classic planimetric composition has mastered foreshortening, so that while the important figures tend to be set on the same plane, the bodies and faces may be angled in a variety of ways. The prototype is Leonardo's Last Supper of 1495–1498.
By contrast, recessional composition is characteristic of baroque art of the 1600s. Here significant action is spread into depth, usually with sharply felt contrasts of scale. A vivid instance is Rubens's Lamentation of 1614, in which the dead Christ's body angles steeply back from the picture plane, and the mourners are arranged in diagonals around him.
Although Bazin makes no reference to Principles of Art History, his account of Welles and Wyler comfortably fits a notion of recessional style.12 Many American films of the 1940s build diagonal compositions throughout, with figures ranged along sharply focused planes. Most vivid is what Wölfflin called “the exaggerated foreground,” which in baroque art “increased the suddenness of perspective reduction.”13 (See Figure 2.) Other features highlighted by Wölfflin are settings that favor depth such as corridors, dark foregrounds, and frames within the frame. All of these techniques can be found as well in 1940s Hollywood, and they lend support to our tendency to label such films “baroque.”
We can, though, go back further. Wölfflin's concept of recession alerts us to another tendency that was not apparent to film historians until recently. Silent‐film historians had decried most pre‐Griffith cinema as essentially theatrical, and to a great extent the Bazinian revision held the same view.14 But we can detect sophisticated pictorial strategies in European and some American filmmaking of the entire period 1906–1920. This is the era of “tableau” filmmaking, with the term signifying at once painting‐like design and a kinship to the stage “pictures” of the period.15
Early film staging displays both recessional and planimetric qualities. A film commonly presented rather shallow interiors and fairly diagonal action outdoors (Figures 3 and 4). Once films began to run ten minutes or more, directors extended their scenes and developed more elaborate indoor sets and more complicated staging. They seldom cut within the scene, but within the extended shot directors explored a great many ways to compose shots that would guide the viewer's attention.
Sometimes the tableau was decidedly planar, with figures stacked in parallel rows. The director might highlight one action through movement or a peekaboo tactic, letting an important face or object be glimpsed in a cranny of the lateral plane (Figure 5). But the tableau style hosted many explorations of recessional shot design as well. A very early and influential example is The Assassination of the Duke de Guise (1908). It was advertised as the height of spectacular theater, but its compositions display a dynamic play between foreground and background. A rather crowded scene makes much use of Wölfflin's “baroque coulisses and encadrements” (Figure 6).16
In the recessional style of the 1910s, sometimes the frame is spacious and the depth is marked by figures with plenty of area in between (Figure 7). At other times the frame is tightly packed, and figures open and fill gaps in the course of the scene (Figures 8 and 9). The images direct our attention to important action through various cues: lighting, centrality of figures in the format, frontality of figures, advance to the foreground, and speed of movement. Figures closer to the camera can mask or reveal figures further away, and this choreography became quite complex in the course of the 1910s (Figures 10 and 11).
At first glance, these manipulations of space within the frame might seem no different from what one would see on the theater stage. But filmmakers of the period were quite aware that the playing space of cinema did not replicate that of the stage.
The proscenium tradition makes theatrical playing space wide and shallow. Staging the action horizontally allows viewers all over the auditorium to grasp the action. By contrast, a film's playing space is created by the optical properties of the lens. Although we seem to be looking into a theatrical box, the area captured by the camera is actually a pyramid tipped at a right angle and tapering at the lens surface. The playing area is narrow but deep. As a result, filmic action must be staged for the only important sightline: that of the camera. Many of the shots illustrated here could not be played on a stage; major areas of action would not be visible from some seats in the auditorium.
The distinctive playing space of cinema made the rich spatial manipulations of the tableau style possible. Ironically, this perspectival capture of moving bodies is as “cinematically specific” as editing was held to be.
Far from being sheer recording of theater, then, the tableau approach constitutes a distinct group style. Individual filmmakers explored it in various ways, but they were working with the same broad schemas of staging. Many filmmakers presented the style in rather pure form, with single‐take scenes interrupted only by intertitles. Other filmmakers coordinated the tableau with editing choices—typically, axial cut‐ins that enlarged a section of the tableau space. This “scene‐insert” approach was characteristic of the style in the later 1910s. Sometimes too we encounter filmmakers who press tableau premises to the limit by striving for very close foregrounds. A scene in Robert Reinert's Nerven (1920) is akin to the big heads we see in Welles and Wyler (Figure 12).17
We might consider the tableau approach a period style, since it was almost completely replaced when the editing‐oriented style, developing at the same time, took over world cinema around 1920. Yet the recessional strategies that were made prominent in the tableau era were amplified in the shot designs of some 1920s and 1930s films, notably Russian ones. And the period's choreographic options became permanent resources of filmmaking. They emerged with particular prominence in long‐take cinema from the 1960s onward.
V. FILM HISTORY AND ART HISTORY: THE PLANIMETRIC STYLE
Cinema displays a history of planimetric composition as well. We find it in marked form in the first decade or so of fictional filmmaking (Figure 13) and in more sophisticated form in the tableau era, as we have seen. A milder sort of plane‐based composition was the default for most dramatic encounters in the classic silent film and the 1930s cinema. Figures face one another on the same plane, in profile or with their heads swiveled a little to yield three‐quarter views (Figure 14). Shot/reverse‐shot editing transposes this face‐off into a series of shots. Now each actor, seen singly, is presented in three‐quarter view, and the answering shot provides an echoing image of the other actor (Figures 15 and 16). But both are usually felt to be on the same plane, especially if the background surfaces are perpendicular to the camera.
The more baroque recessional compositions that became popular in the 1940s and 1950s were, across international cinema, the major alternative to mild planimetric layouts. And although the depth of field necessary for the big foregrounds of Welles, Wyler, and others was difficult to achieve in color and widescreen filmmaking, some filmmakers found ways to get it (Figure 17). What was unexpected was the emergence of a more severe planimetric look from the 1960s onward.
Directors began setting up the camera perpendicular to the background surface. Figures stood in distinct rows facing the viewer or turned ninety degrees (Figures 18 and 19). The whole array was lit flatly, to downplay depth. Just as the other stylistic trends surfaced in many countries, this faux‐naïf planimetric look became in ensuing decades one option of international cinema. It could be used for groups or for individual confrontations (Figures 20 and 21). In the 1990s, Takeshi Kitano made it central to his films (Figure 22), and by 2014 Wes Anderson had established a variant of it
Wölfflin argued that the sophisticated planimetric style of the Cinquecento differed from the “primitive” variant because it had mastered the technical innovations of foreshortening. Likewise, cinema's new planimetric look could succeed only after editing permitted what I have called compass‐point cutting: 180‐degree reversals that showed what the characters were looking at (usually other characters facing the camera). (See Figures 23 and 24.)
VI. EXPLANATIONS
Wölfflin's frame of reference, then, facilitates our discovering and describing some important changes in cinematic representation. But how to explain them? We need not mobilize some explanations that Wölfflin would find congenial, such as the appeal to national character. Nor need we embrace his other major causal mechanism, the idea that “vision itself has a history.”18 The point is not quite as stark as it appears: he indicates that a better term than “modes of vision” would be “modes of imagination.”19 Yet given the coexistence of so many film styles across just a few decades, this epochal concept is not a plausible option for film research.
More plausible explanatory frameworks have been proposed by another pictorial poetician, E. H. Gombrich. While Wölfflin leans toward methodological holism, Gombrich posits formal development as explicable largely through methodological individualism. “The artist is always faced with the problem: ‘What is there for me to do?’”20 Further, Gombrich endows the concept of artistic schemas with a more tangible existence than Wölfflin does. They exist less in the artists’ heads than in the pattern books and traditional pictorial solutions that novices encounter during their training. Most vigorously, Gombrich resists national character and “race” as causal factors, focusing instead on fashion, competition, and the social functions of art within a given artistic community.21
I find a framework conceived in terms of problem and solution and schema and revision more suggestive when studying an art dependent on a mass market, a tightly knit community of creators, and well‐established norms of sound, image, and storytelling. Accordingly, I think we should look to causes of a Gombrichian sort: production conditions, influences, fashion, competition, and the urge to distinguish one's work. For example, the complex staging of the tableau style may have begun from theatrical and pictorial models, but they were transformed through trial and error into something that fitted the optical perspective of the camera. Filmmakers could borrow pictorial schemas from perspective painting and shape them to the constraints of movement and camera position.
In later years, the rise of extreme planimetric imagery arose, I believe, from a chain of factors. A new reliance on location shooting in the 1960s pushed filmmakers to use long lenses, which in turn yielded images with fewer depth cues. The “flattening” of the image, along with the new widescreen ratios, encouraged some filmmakers, especially those inclined to long takes, to stage action frontally and laterally. There was an expressive advantage as well. The schema could further the “dedramatization” characteristic of art cinema, and it presented a sharp alternative to the brash dynamism of Hollywood.
Yet we should remember Wölfflin's suggestion that these schemas are primarily decorative in nature, without inherent expressive qualities. The tableau style was just as suited to comedy as to melodrama, and the faux‐naïf planimetric look was recruited for both low‐affect comedy and high drama in mainstream American entertainments. In all, Wölfflin has left us a legacy of midrange concepts that help specify and analyze pictorial traditions in moviemaking. Yet his work also reminds us just how complex a phenomenon visual style is, and how stubbornly it resists explanations.
Footnotes
1.The purported contrasts between German and Italian national character are developed in detail in Heinrich Wölfflin, The Sense of Form in Art: A Comparative Psychological Study, trans. Alice Muehsam and Norma A. Shetan (New York: Chelsea House, 1958). A typical remark is that the painterly tendency (as opposed to the linear one) “runs in the veins of the Germanic race.” Principles of Art History: The Problem of the Development of Style in Later Art, 7th edition [1923], trans. M. D. Hottinger (New York: Dover, 1950), p. 67.
2.Erwin Panofsky, “Style and Medium in the Motion Pictures,” in Three Essays on Style, ed. Irving Lavin (MIT Press, 1995), pp. 91–125, at p. 108.
3.See John Frazer, Artificially Arranged Scenes: The Films of George Méliès (Boston: G. K. Hall, 1979).
4.Heinrich Wolfflin, Classic Art: An Introduction to the Italian Renaissance, trans. Peter and Linda Murray (London: Phaidon, 1952), p. 5.
5.The argument is set out most fully in André Bazin, “The Evolution of the Language of Cinema,” in What Is Cinema? vol. I, ed. and trans. Hugh Gray (University of California Press, 1967), pp. 23–40.
6.Bazin, “Evolution of the Language of Cinema,” p. 35.
7.Bazin, “Evolution of the Language of Cinema,” p. 38.
8.Examples are Lewis Jacobs's analyses of crosscutting in sequences of Griffith's Birth of a Nation and Intolerance in The Rise of the American Film: A Critical History (Columbia University Teachers College Press, 1968), pp. 180–198.
9.The relevant argument is found in Wölfflin, Principles of Art History, chap. II.
10.See Wölfflin, Principles of Art History, pp. viii, 14, 233; and Heinrich Wölfflin, “‘Kunstgeschichtliche Grundbegriffe’: eine Revision,” in Gedanken zur Kunstgeshichte (Basel: Schwabe, 1941), pp. 18–24, at p. 23.
11.Wölfflin, The Sense of Form in Art, p. 4.
12.Principles appeared in French translation in 1952, some years after Bazin's initial discussions of deep focus. It is likely that Bazin was sensitized to the new style by 1940s press material signed by Wyler, Welles, and cinematographer Gregg Toland. See David Bordwell, On the History of Film Style (Harvard University Press, 1997), pp. 56–57.
13.Wölfflin, Principles of Art History, p. 84.
14.True, Bazin claimed to find a “primitive” profondeur de champ in early cinema, but his example, an undated image from the Onésime series, is striking exactly because of its similarity to the aggressive foregrounds of Welles. See Bordwell, On the History of Film Style, pp. 62–63.
15.Major investigations of tableau cinema are Ben Brewster, “Deep Staging in French Films 1900–1914,” in Early Cinema: Space Frame Narrative, ed. Thomas Elsaesser (London: British Film Institute, 1990), pp. 45–55; Yuri Tsivian, “Portraits, Mirrors, Death: Death: On Some Decadent Clichés in Early Russian Films,” Iris 14–15 (1992): 70–78; Tom Gunning, “Notes and Queries about the Year 1913 and Film Style: National Styles and Deep Staging,” 1895, special issue: “L'année l9l3 en France” (1993): 195–204; Ben Brewster and Lea Jacobs, Theatre to Cinema (Oxford University Press, 1997), pp. 164–187. The account of the tableau style and planimetric staging that follows is drawn from Bordwell, On the History of Film Style, pp. 158–207, and David Bordwell, Figures Traced in Light: On Cinematic Staging (University of California Press, 2005). See also entries on my website, www.davidbordwell.net.
16.Wölfflin, Principles of Art History, p. 86.
17.For more on Reinert, see my essay “Taking Things to Extremes: Hallucinations Courtesy of Robert Reinert,” in Poetics of Cinema (New York: Routledge, 2008), pp. 281–326.
18.Wölfflin, Principles of Art History, p. 11.
19.Foreword to the sixth edition, Principles of Art History, p. vii. It seems clear that “modes of vision” are not purely optical phenomena. Wölfflin explains that the artist internalizes “what choice of formal possibilities the epoch had at its disposal.” This fits the famous claim that immediately precedes the statement about vision's history: “Every artist finds certain visual possibilities before him, to which he is bound. Not everything is possible at all times” (Principles of Art History, p. 11). In the 1933 essay “‘Kunstgeschichtliche Grundbegriffe’: eine Revision,” Wölfflin explains that he used the term “mode of vision” in a sense analogous to speaking of “the artist's ‘eye’ and his ‘vision,’ by which we mean the manner in which objects take form in an internal representation” (Gedanken zur Kunstgeshichte, pp. 19–20).
20.Quoted in Ernst Gombrich and Didier Eribon, Looking for Answers: Conversations on Art and Science (New York: Abrams, 1993), p. 168.
21.This duality is not quite as sharp as I am making it out. Wölfflin, for instance, suggests that one major source of the artist's “mode of beholding” is prior art, “the effect of picture upon picture”—a tenet of Gombrich's thinking as well. See Wölfflin, Principles of Art History, pp. 229–230; E. H. Gombrich, Art and Illusion: A Study in the Psychology of Pictorial Representation (Princeton University Press, 1969), pp. 63–90.
