I actually made a mistake earlier when I said deaf people thinking in sign language would not be thinking in verbal language. Written text, oral communications, Asian symbols and sign language will all be classified as 'verbal'. I think the primary characteristic is that a set of grammatical rules and abstracted labels are used together to describe complex systems. On the other hand thinking in visual models that don't follow a specific set of rules or use a specific set of labels, and which don't abstract as much, would be considered non-verbal thinking, not because it is visual but because it isn't constructed from a set of abstracted component parts and organized by a set of rules. If I write 'The dog jumps over the fence" it is verbal, the same if I write it in Chinese symbols. It is still verbal if I say it orally, or if I sign it with my hands or imagine it being signed. But if I visualize a dog jumping over a fence then I think that is no longer verbal thinking but rather is visuospatial thinking. I am not an expert regarding such things but the types of thinking we are discussing are defined in psychology. So to answer your original question, yes is is entirely possible to think without language, most humans think in both pictures and in language. Some humans are almost only able to think in either pictures or words (usually they are on the Autism spectrum). I originally took your question to mean 'understanding without actualizing a thought' or something like that anyway. 'Knowing without expressing' perhaps. Classical Autistic Thinking Style (think in pictures, limited ability to think in words, need to translate words to pictures): http://autisticsite.com/autistic-children-thinking-in-pictures-instead-of-words I cannot find a web article (only .pdf files) discussing types of Autism where thinking is predominantly in words with visuospatial information needing to be translated, but it is manifested in many with Aspergers syndrome. Additionally, it is a hallmark of NLD, which many mental health professionals consider to be a form of Aspergers. Such people would, for example, have an ability to understand the word 'happy' and link it to the emotional state, but an inability to link a picture of a smiling face to the emotional state. This is in contrast to classically autistic people, who would have an ability to link a picture of a smiling face to the emotion of happiness, but have an inability to link the word 'happy' to the emotion. Therefor a classically autistic person may think 'music makes me happy' by imagining a stereo and then a smiling face, whereas someone with an inability to think in pictures would think 'music makes me happy' (either in written text, sign language, oral words, Asian symbols, etc). Inability to think visuospatially is especially detrimental to navigational tasks. Most people navigate spatially by remembering the layout of the land, landmarks and their orientation in space. People who have an inability to think spatially must use vastly inferior navigational techniques. For example, somebody with the ability to think visuospatially may go to a friends house a ways down the street from their own, and when it is time for them to go home they can visually recall how to return to their home. For somebody who is only capable of thinking in words this is a much more challenging task. The person who can only think in words needs to resort to inefficient techniques, such as counting the number of houses between their own home and the home of their friend. The more complex the navigational task is the more disadvantaged the person who can only think in words is. The visuospatial thinker can visually recall how to get to places ("Well, this looks familiar and I can use these visual landmarks to determine where I currently am in space, and from that I can determine the relationship between myself and my destination") but the verbal thinker must encode the entire route verbally ("first I walk down the street towards the big tree, then I keep going until I find main street, then I turn left on main street and walk ten houses down"). When the ability to verbally encode is controlled for, people with severe hippocampial damage are incapable of finding a fixed location even in a small area, even if they have already been shown where it is, and even when there are multiple landmarks to help them orient themselves. They always move around randomly trying to find the location. This is called landmark orientation. Some people also have deficits in geometric orientation. When placed in a rectangular room and shown a corner, they are then spun around and disoriented. When asked to find the corner originally pointed to, they always randomly guess. Somebody with an ability to think visuospatially would visualize the corner that was pointed to and be able to make note that the wall to the left/right was either long or short (thus having a 1/2 probability of being correct, rather than 1/4). This sort of geometric orientation is automatic and natural for most people, using a verbal strategy to solve this problem requires forethought and an effort to consciously encode the information verbally. And as far as 'understanding without actualizing a thought' or 'Knowing without expressing', I can think of a good example. If I hear an explosion, I have an understanding that something exploded prior to thinking in my mind 'something exploded!' or visualizing something exploding. So I have the original thought prior to expressing it either verbally or visually, but then my brain automatically pipes the thought into either a verbalization (I think predominately in words) or a visualization (I sometimes think visually, but it is much rarer for me). I cannot really manipulate the understanding without verbalization or visualization, but I can still come to the understanding that there was an explosion prior to actualizing the understanding in a manipulable form in my mind. Indeed I must have the understanding prior to piping it into either verbalization or visualization.