What is filling-in? It is the phenomenon in which an empty region of visual space appears to be filled with the color, brightness or texture of its surround. The brain is capable of filling-in the blind spot, borders, surfaces and objects. To explain each case there are two main types of theory, suggesting that filling-in is either isomorphic or symbolic.
什么是填充？它是指视觉空间的空白区域似乎被周围的颜色、亮度或纹理填充的现象。大脑能够填充盲点、边界、表面和物体。为了解释每种情况，有两种主要的理论认为填充是同构的或符号的。

The natural blind spot is a retinal region devoid of photoreceptors, where the head of the optic nerve joins the retina (Figure 1A). Close your left eye and extend your right arm straight in front of you. Spread your fingers wide, palm down, and gaze at your thumbnail. Now wiggle your little finger. It will fall on your blind spot and you will be unable to see it! The blind spot is quite large, about $6^{\circ }$$6^{\circ }$6^(@)6^{\circ} in diameter - large enough to hold a dozen moons side by side. So it makes a big hole in the visual field - yet we do not see a hole. It is true that the retina of the other eye covers this area, but even with the other eye closed the area of the blind spot looks ‘filledin’ with the color or texture of the surround. If the surround is red or black or striped, then the region of the blind spot appears to be filled in with red or black or with stripes. Also, traumatic damage to the retina, or the slow degeneration caused by glaucoma, can result in a blind spot or scotoma: a scotoma is also perceptually filled-in, so effectively that the patient is often not aware of it, which can lead to undesirable delay in seeking medical treatment.
自然盲点是视网膜上没有光感受器的区域，也就是视神经头与视网膜连接的地方（图 1A）。闭上左眼，将右臂伸直放在面前。手指张开，掌心向下，注视着拇指指甲。现在摆动你的小拇指。它将落在你的盲点上，你将无法看到它！盲点相当大，直径约 $6^{\circ }$$6^{\circ }$6^(@)6^{\circ} --足以并排容纳十几个月亮。因此，它在视野中形成了一个大洞--但我们却看不到一个洞。的确，另一只眼睛的视网膜覆盖了这一区域，但即使闭上另一只眼睛，盲点区域看起来也会被周围的颜色或纹理 "填满"。如果周围是红色、黑色或条纹状，那么盲点区域看起来就是被红色、黑色或条纹填满的。此外，视网膜的外伤性损伤或青光眼引起的缓慢变性也会导致盲点或视网膜疤痕：视网膜疤痕也会在知觉上被填满，以至于患者往往意识不到它的存在，这可能会导致患者延误就医。

If the eyes fixate a point very steadily, small objects in peripheral vision often fade out from view and disappear. This was first reported by Troxler in 1804. This process is accelerated if the edges of the object are blurred, and is even faster and more complete if
如果眼睛非常稳定地盯住一个点，周边视野中的小物体往往会从视野中淡出并消失。特罗克斯勒在 1804 年首次报道了这一现象。如果物体的边缘模糊不清，这一过程就会加快，而如果
these edges are stabilized against the jitter caused by small eye movements by means of special lenses. If a red disk is surrounded by a green annulus, and the border of the red disk, but not of the green annulus, is retinally stabilized, then the red disk is gradually filled-in perceptually with the green color of the annulus, and the whole field looks green. Although the red disk is still present in the field, it becomes invisible.
这些边缘通过特殊的镜片来稳定，以防止眼球微小运动造成的抖动。如果一个红色圆盘被一个绿色环状体包围，而红色圆盘的边缘（而不是绿色环状体的边缘）在视网膜上被稳定，那么红色圆盘在知觉上就会逐渐被绿色环状体的绿色所填满，整个视野看起来就是绿色的。虽然红色圆盘仍然存在于视野中，但它变得不可见了。

Consider a large field filled with twinkling noise, like the snow on a detuned TV set, with a fixation point at its center and a small embedded window that is viewed peripherally. During prolonged fixation on the center, the window and its contents gradually disappear. If the window is filled with random dots that drift to the left (with the window’s edges remaining stationary), or simply with grey, then the window fades out from view within 5-20 seconds. The smaller and more peripheral the target, the faster it disappears. When the display is abruptly replaced by a uniform grey field,
考虑一个充满闪烁噪声的大场，就像失谐电视机上的雪花一样，其中心有一个固定点，外围有一个嵌入式小窗口。在长时间盯住中心点的过程中，窗口及其内容会逐渐消失。如果窗口内填充的是向左漂移的随机点（窗口边缘保持不动），或者仅仅是灰色，那么窗口会在 5-20 秒内从视线中消失。目标越小、越边缘化，消失得就越快。当显示屏突然被统一的灰色区域取代时、
the area of the window appears to be filled with an aftereffect: respectively, a motion aftereffect apparently drifting to the right, or an aftereffect of apparent twinkle.
窗口区域似乎被某种效果填满：分别是明显向右漂移的运动效果，或明显闪烁的效果。

This is not simply adaptation to luminance edges of the window they are constantly refreshed by the twinkling dots that alter the contrast of the edges. Even a flickering spot on a plain grey surround gradually fades out, and to keep it visible the amplitude of flicker must be increased steadily over time. Wiggly lines gradually look straighter. All these cases show a gradual loss of visual information in the peripheral visual field, occurring in two stages; at first the borders of the patch act like a dam that initially erodes gradually over time; and then the surround floods in rapidly like water over the dam, actively filling-in the patch until it is invisible. This competition between two surfaces, in which the surround surface gradually wins, may not be the same process as filling-in of the blind spot, which is instantaneous; nobody knows.
这不仅仅是对窗口亮度边缘的适应，闪烁的光点会不断刷新窗口亮度，从而改变边缘的对比度。即使是在纯灰色环境中闪烁的光点也会逐渐消失，为了保持其可见度，闪烁的幅度必须随着时间的推移不断增加。扭曲的线条逐渐变得平直。所有这些情况都表明，外围视野中的视觉信息会逐渐消失，并分为两个阶段：首先，斑块的边界就像一道堤坝，最初会随着时间的推移逐渐被侵蚀；然后，周围的环境就像大坝上的水一样迅速涌入，主动填满斑块，直到看不见为止。这种两个表面之间的竞争，即环绕表面逐渐获胜的过程，可能与盲点的填充过程不同，后者是瞬间完成的，无人知晓。

B

C

Current Biology 当前生物学

Figure 1. Filling-in of the blind spot and of surfaces.
图 1.填充盲点和表面。
(A) The blind spot. Close your left eye, gaze at the cross, and move the page toward you. At some point the black spot will disappear because it lands on your retinal blind spot. However, the red and green stripes perceptually fill into the blind space. (B) Neon spreading. The thin red lines are perceptually filled in to form an illusory pink annulus. Both rings are the same red, but they average together with their white or black backgrounds to look light or dark pink. © Pinna’s water color illusion. The colored lines appear to fill-in and tinge the entire regions with color.
(A) 盲点。闭上左眼，注视十字架，然后将页面向您移动。在某一时刻，黑点会消失，因为它位于视网膜盲点上。然而，红色和绿色条纹会在知觉上填满盲区。(B) 霓虹灯扩散。红色细线在知觉上被填满，形成一个虚幻的粉红色环。两个环都是相同的红色，但它们与白色或黑色背景平均在一起，看上去呈现出浅粉色或深粉色。平纳的水色幻觉。彩色线条似乎填满了整个区域，并给整个区域染上了颜色。

intersections of a mesh of thin black lines are colored, say red, then each red intersection seems to expand into a pink disk. This bleeding of color, which resembles diffusion, is called neon spreading. It is strongest if the grey surround has the same luminance as the red color, and is weakened if a thin white gap (not shown) is left between each red cross and the black lines aligned with it. In Figure 1B, the red crosses seem to amalgamate and form a complete pink annulus.
如果在黑色细线网格的交叉点上涂上颜色，比如红色，那么每个红色交叉点似乎都会扩展成一个粉红色的圆盘。这种类似于扩散的渗色现象被称为霓虹扩散。如果灰色环绕的亮度与红色相同，则这种渗色效果最强；如果在每个红色交叉点和与其对齐的黑线之间留有一条细长的白色空隙（未显示），则渗色效果会减弱。在图 1B 中，红色十字似乎合并在一起，形成一个完整的粉红色环。

In Pinna’s water color illusion, a shape is drawn with a wiggly black outline (Figure 1C). Running along just inside the black is a wiggly colored line, say yellow, which makes the whole shape appear to be tinged with yellow. This effect was well known to early cartographers, who tinted coastlines blue in their maps to suggest the blue of the ocean. Watercolor and neon colors spread, provided they are equiluminous with their surround, in a visual process analogous to physical diffusion, until they encounter a luminance contour that acts as a barrier and prevents further spreading.
在 Pinna 的水彩幻觉中，一个形状被画上了一条摇摆的黑色轮廓线（图 1C）。在黑色轮廓的内侧，有一条晃动的彩色线条，比如黄色，这使得整个形状看起来都染上了黄色。早期的制图师对这种效果非常熟悉，他们在地图中将海岸线染成蓝色，以暗示海洋的蓝色。水彩和霓虹灯的颜色在与周围亮度相等的情况下会扩散，其视觉过程类似于物理扩散，直到遇到亮度轮廓线作为屏障，阻止颜色进一步扩散。

completion? Objects in the visual world are often partly covered or partly visible, so we need to be able to fill in the gaps and perceive the whole object. If a target is partly covered by an occluder, ‘amodal completion’ refers to completing the perception of the hidden parts that lack any visible attributes (color, texture and so on). On the other hand, Kanisza’s triangle (Figure 2A) produces ‘modal completion’ in front of the supposedly occluded pacmen. The illusory triangle looks slightly brighter than the surround, and in front of it, though there are no corresponding luminance edges present. These illusory contours are modal, being perceptually salient and appearing to belong to the triangular figure rather than the ground.
完成？视觉世界中的物体通常会被部分遮挡或部分可见，因此我们需要能够填补空白，感知整个物体。如果目标物被遮挡物部分遮挡，"模态完成 "指的是完成对缺乏任何可见属性（颜色、纹理等）的隐藏部分的感知。另一方面，卡尼斯扎的三角形（图 2A）在假定的遮挡物前产生了 "模态完成"。虽然没有相应的亮度边缘，但虚幻的三角形看起来比周围稍亮，而且就在它的前面。这些虚幻的轮廓是模态的，具有知觉上的显著性，看起来属于三角形图形而非地面。

The illustrations in Figure 2B-D bring out the subjective difference between modal and amodal completion. By crossing one’s eyes in free fusion it is possible to combine $C$$C$CC and $D$$D$DD. The disparities are arranged so that a grey triangle appears to lie in depth behind
图 2B-D 中的插图显示了模态完成和非模态完成之间的主观差异。在自由融合时，通过交叉双眼，可以将 $C$$C$CC 和 $D$$D$DD 结合起来。这样，一个灰色的三角形就会出现在$C$$C$CC和$D$$D$DD的后面。

Figure 2. Modal and amodal completion.
图 2.模态和非模态完成。
(A) Modal completion: Kanisza’s illusory triangle. This appears to occlude the pacmen and thin black lines. It seems to float in front and be slightly brighter than the background. (B-D) If you cross your eyes and fuse $C$$C$CC and $D$$D$DD together binocularly, a grey triangle is visible lying in depth behind the three portholes. This is amodal completion: the hidden part of the triangle is not assigned a particular color. But if you fuse $B$$B$BB and $C$$C$CC together, the grey triangle now floats in depth above the portholes - and modal completion fills-in its ghostly grey center that floats above the surround.
(A) 模态完成：卡尼萨的虚幻三角形。它似乎遮住了小人和黑色细线。它似乎漂浮在前面，比背景稍亮。(B-D) 如果交叉双眼，用双目将 $C$$C$CC 和 $D$$D$DD 融合在一起，就会看到一个灰色三角形位于三个舷窗后面的深处。这就是模态完成：三角形的隐藏部分没有被赋予特定的颜色。但是，如果将 $B$$B$BB 和 $C$$C$CC 融合在一起，灰色三角形现在就会漂浮在舷窗上方的深处--模态补全填充了漂浮在周围上方的灰色幽灵中心。
three triangular portholes. The grey triangle is amodally complete - it looks like a single triangle, not like three unrelated segments - but the brightness of its hidden parts is not perceptually represented. If one binocularly combines $B$$B$BB and C , however, one sees in stereo the grey triangle lying in depth in front of the three portholes. Now the ghostly middle of the grey triangle is perceptually - modally - filled-in, in front of the black surround.
三个三角形舷窗。灰色三角形在模态上是完整的--它看起来像一个三角形，而不是三个互不相关的部分--但其隐藏部分的亮度在知觉上并没有表现出来。然而，如果将 $B$$B$BB 和 C 结合在一起，就会在立体声中看到灰色三角形位于三个舷窗前方的深度。现在，在黑色环绕物的前面，灰色三角形幽灵般的中间部分在感知上--模态上--被填充了。

What causes filling-in? Edges and contours are important in the filling-in of colors, brightness and textures into surfaces. There are two theories of how this might be done: isomorphic or neural filling-in, and symbolic filling-in.
填充的原因是什么？边缘和轮廓对于色彩、亮度和纹理的表面填充非常重要。关于如何实现填充，有两种理论：同构填充或神经填充，以及符号填充。
According to the isomorphic theory, a surface to be filled-in is represented in the brain by a twodimensional array of neurons, and these actually fire under the influence of excitation spreading in from the edges. It is suggested that color signals spread in all directions until
根据同构理论，要填充的表面在大脑中由一个二维神经元阵列表示，这些神经元实际上是在从边缘扩散进来的兴奋的影响下发射的。有人认为，颜色信号会向各个方向传播，直到
they are stopped by a luminance border that acts as a barrier. This process is thought to be analogous to physical diffusion.
它们会被作为屏障的亮度边界挡住。这一过程被认为类似于物理扩散。

By the alternative theory, symbolic filling-in, there is no spread of activity within the surface. Instead, visual properties of the surround, such as texture, contrast polarity and color, are tagged and applied to the enclosed surface. This process is sparse and economical and resembles the vector graphics method of representation. It is not clear how this would be implemented in the nervous system. On some minimalist version of symbolic fillingin, little implementation is needed; thus, the retinal blind spot might be ignored, as the gap behind our heads is ignored, since after all there are no neurons in the brain devoted to seeing the blind spot.
根据另一种理论，即符号填充理论，表面内的活动不会扩散。相反，周围的视觉特性，如纹理、对比极性和颜色，会被标记并应用到被包围的表面上。这一过程既稀疏又经济，类似于矢量图形的表示方法。目前还不清楚如何在神经系统中实现这一功能。因此，视网膜盲点可能会被忽略，就像我们脑后的缝隙被忽略一样，因为大脑中毕竟没有专门用于观察盲点的神经元。

What is the evidence for isomorphic filling-in? Strong support comes from motion aftereffects around the blind spot. It is known that, if one first inspects a
同构填充的证据是什么？盲点周围的运动后效提供了强有力的支持。众所周知，如果首先检查一个
striped moving pattern, then when the motion is stopped the stripes appear to move back in the opposite direction. This motion aftereffect can be built up to one eye and elicited from the other eye. Murakami aimed an annular stimulus containing drifting stripes at the blind spot of the right eye. Filling-in made this look like a complete striped disk. He then tested with a small striped disk, smaller than the hole in the annulus, and viewed by the corresponding intact retina of the left eye. Observers reported a motion aftereffect. This suggests that the motion was actively filled into the blind spot of the right eye.
当运动停止时，条纹会向相反的方向移动。这种运动后效可以通过一只眼睛建立，并从另一只眼睛激发。村上将含有漂移条纹的环形刺激物对准右眼盲点。填充物使其看起来像一个完整的条纹圆盘。然后，他用一个小的条纹圆盘进行测试，该圆盘比环状刺激物上的洞小，由左眼相应的完整视网膜观看。观察者报告出现了运动后效。这表明运动被主动填充到了右眼的盲点。

Further support comes from fillingin of a large disk. A large white disk was flashed up on a black surround, followed 50-100 milliseconds later by a white outline circle of about half the diameter. Observers reported that the interior of the outline circle looked black. If the circle was smaller, the optimal time gap was longer. This suggests that a brightness signal propagates inward from the circumference at an estimated speed of $110-{150}^{\circ }$$110-{150}^{\circ }$110-150^(@)110-150^{\circ} of visual angle per second, and is arrested by the barrier of the outline circle. This is supported by experiments on simultaneous contrast. A static grey disk appears to brighten and dim when its surround respectively dims and brightens, but this process breaks down at repetition rates exceeding 2.5 Hz , suggesting also a spreading process at a finite speed. Similar effects have been shown for the filling-in of texture.
一个大圆盘的填充提供了进一步的支持。一个巨大的白色圆盘在黑色环绕中闪现，50-100 毫秒后，一个直径约为白色圆盘一半的白色轮廓圆出现。观察者报告说，轮廓圆的内部看起来是黑色的。如果圆圈较小，则最佳时间间隔较长。这表明，亮度信号以每秒 $110-{150}^{\circ }$$110-{150}^{\circ }$110-150^(@)110-150^{\circ} 视角的估计速度从圆周向内传播，并被轮廓圆的屏障阻挡。这一点得到了同步对比实验的支持。当一个静态灰色圆盘的周围分别变暗和变亮时，圆盘就会变亮和变暗，但当重复频率超过 2.5 Hz 时，这一过程就会中断，这也表明了一个有限速度的扩散过程。纹理填充也有类似的效果。

Dennett, D.C. (1991). Consciousness explained. Boston, Toronto, London: Little, Brown.
Dennett, D.C. (1991).Consciousness explained.波士顿、多伦多、伦敦：Little, Brown.
Komatsu, H. (2006). The neural mechanisms of filling in. Nat. Rev. Neurosci. 7, 220-231.
Komatsu, H. (2006).填充的神经机制》。Nat.Rev. Neurosci.7, 220-231.
Motoyoshi, I. (1999). Texture filling-in and texture segregation revealed by transient masking. Vision Res. 39, 1285-1291.
Motoyoshi, I. (1999).瞬时遮蔽揭示的纹理填充和纹理分离。Vision Res. 39, 1285-1291.
Murakami, I. (1995). Motion aftereffect after monocular adaptation to filled-in motion at the blind spot. Vision Res. 35, 1041-1045.
Murakami, I. (1995).单眼适应盲点填充运动后的运动后效。Vision Res. 35, 1041-1045.
Paradiso, M.A., and Nakayama, K. (1991). Brightness perception and filling-in. Vision Res. 31, 1221-1236.
Paradiso, M.A. 和 Nakayama, K. (1991)。亮度感知与填充。Vision Res. 31, 1221-1236.
Pessoa, L., and De Weerd, P. (2003). Filling-In: From Perceptual Completion to Cortical Reorganization. (Oxford: Oxford University Press).
Pessoa, L. 和 De Weerd, P. (2003)。Filling-In：From Perceptual Completion to Cortical Reorganization.(牛津：牛津大学出版社）。
Pessoa, L., Thompson, E., and Noe, A. (1998). Finding out about filling-in: A guide to perceptual completion for visual science and the philosophy of perception. Behavi. Brain Sci. 21, 723-802.
Pessoa, L., Thompson, E., and Noe, A. (1998).发现填充：视觉科学与知觉哲学的知觉补全指南》。Behavi.21, 723-802.

Department of Psychology, UC San Diego, 9500 Gilman Drive, La Jolla, CA 92093, USA.
美国加州大学圣地亚哥分校心理学系，地址：9500 Gilman Drive, La Jolla, CA 92093。
E-mail: sanstis@ucsd.edu
电子邮件：sanstis@ucsd.edu

Tudor domain was first identified as a segment of approximately 60 amino acids that is present in 11 repeated units in the Drosophila protein of the same name. Drosophila tudor was first identified genetically, in a large-scale screen for maternaleffect lethal mutations that affected embryonic development. Several complementation groups of such mutations were identified in which homozygous females produced embryos that failed to specify primordial germ cells, and these were named after extinct European royal families (tudor, vasa, valois, and staufen). Since that time, over 200 Tudor-domain containing proteins have been identified from essentially
Tudor 结构域最初被鉴定为果蝇同名蛋白质中 11 个重复单元中约 60 个氨基酸的片段。果蝇 Tudor 的首次遗传鉴定是在大规模筛选影响胚胎发育的母系效应致死突变时进行的。在这种突变的几个互补群中，发现同卵雌果蝇产生的胚胎不能指定原始生殖细胞，这些互补群以已灭绝的欧洲王室命名（都铎、瓦萨、瓦卢瓦和斯托芬）。从那时起，已经从基本上所有的欧洲皇室家族中鉴定出了 200 多种含有都铎结构域的蛋白质。
all varieties of eukaryotes, including plants, animals, and fungi, but not from prokaryotes. Tudor domains are related to Chromo, MBT, PWWP, and Agenet-like domains, which are implicated in chromatin binding. The core Tudor domain forms a $\beta$$\beta$beta\beta-barrel like core structure that contains four short $\beta$$\beta$beta\beta-strands followed by an $\alpha$$\alpha$alpha\alpha-helical region (Figure 1). In different types of Tudor-domain containing proteins, the core Tudor domain or domains can be flanked on the amino-terminal side with other conserved motifs.
所有真核生物，包括植物、动物和真菌，但不包括原核生物。Tudor结构域与Chromo、MBT、PWWP和Agenet-like结构域有关，这些结构域与染色质结合有关。Tudor 核心结构域形成一个 $\beta$$\beta$beta\beta 桶状核心结构，其中包含四条短 $\beta$$\beta$beta\beta 链，其后是一个 $\alpha$$\alpha$alpha\alpha 螺旋区域（图 1）。在不同类型的含都铎结构域的蛋白质中，核心都铎结构域或多个结构域的氨基末端可与其他保守基团相连。

What is the function of the Tudor domain? Four types of Tudor domains can be distinguished based on their flanking sequences. The original germ-line type Tudor domain binds to proteins with dimethylated arginine or lysine residues. Work in mammals and Drosophila is consistent with a model that arginine methylation of Piwi-type proteins, and their consequent binding to Tudor proteins, is necessary to
都铎结构域的功能是什么？根据其侧翼序列，可将 Tudor 结构域分为四种类型。最初的种系型 Tudor 结构域与具有二甲基化精氨酸或赖氨酸残基的蛋白质结合。哺乳动物和果蝇的研究结果与以下模型一致，即 Piwi 型蛋白的精氨酸甲基化及其与 Tudor 蛋白的结合对于

Current Biology 当前生物学

Figure 1. The Tudor domain.
图 1.都铎领地
Schematic representation of the structure of the two Tudor domains of the human fragile $X$$X$XX mental retardation protein FXR2 (PDB: 3H8Z). The structure was generated by the Structural Genomics Consortium (www.thesgc.org) and placed in the public domain. Each Tudor domain contains four $\beta$$\beta$beta\beta-strands (depicted by broad colored arrows) that form a barrel structure.
人类脆性 $X$$X$XX 精神发育迟滞蛋白FXR2（PDB：3H8Z）的两个Tudor结构域的结构示意图。该结构由结构基因组学联合会（www.thesgc.org）生成并置于公共领域。每个都铎结构域包含四条 $\beta$$\beta$beta\beta 链（用宽颜色箭头表示），形成桶状结构。