The brain is an organ that locates its experiences elsewhere. Every time we "see" something, the experience is a neurological process happening at the end of a causal chain of events starting with (according to our current model) light reflected from objects, passing through our eye lenses, being absorbed in the retial cells, initiating a cascade of nerve impulses, etc., and finally activating memory structures and active neurological experiences that we personally experience as happening "out there" instead of "in here". The brain "projected" the visual experience "onto" the external world, but that experience is physically happening not only NOT "out there", but it is happening delayed in time by large fracions of a second after the causative event.
This is "instantaneous" physiological projection.
When we quickly look at a familiar scene, our brain brings large portions of the image from our memory to construct the scene, so we are "projecting" a significantly large portion of the visual experience from memory rather than from new visual system abstracting. This has been tested, and it's why my signature says, "It's not that seeing is believing, believing is seeing, and we're much better at believing than we are at seeing." Eye witness reports are notoriously unreliable, because the "images" are reconstructed using presumptions that are sometimes false. The witness "believes" what they reconstructed even when it can be proven that they are wrong.
In psychological terms, we attribute to the motivations of others our own experience and motivations.
The neurotic also deny their own motives by "projecting" them on others.
"Projection" happens continuously and all the time, and with varying degrees of valitity or correspondence with "reality". Most of the time we act as if they really are there, but we also have corrective feedback when we discover we were wrong. Our projections are continually being tested.
Do not think of "projection" as being an instance of incorrectly getting the "map" wrong. Projection is always happening at various levels, and like abstraction, it has the potential of getting it wrong at every level. The lowest levels, however, such as direct sight, are usually so good, that we confused the map with the territory - we identified the projection with "reality" - we failed to distinguish the "thing" from our "object" level experience. The great philosophers - going back as far as Plato and his metaphor of the cave - understood the distinction in varying degrees of specificity and using various metaphors. See Heraclitus? or Xenophanes? for one of the earliest formulations of this view.
But the "common man" and even most "educated persons" still have no idea of this mechanism, except in the lay psychological perspective promoted by pop psychology.
'Projection' is the correct word for the way you are using it.
A psychiatric technical definition - projection: A defense mechanism, operating unconsciously, in which what is emotionally unacceptable in the self is unconsciously rejected and attributed (projected) to others.
Example. The spouse who was chronically unfaithful in the first marriage and is continually suspicious of the new spouse.
More info: "Briefly the general theses are: 1, a tendency in the individual person to view others and the environment in his own likeness—in an 'automorphic' way; 2, the automorphic view of the environment is the mechanism of projection; 3, the process of projection takes place outside of awareness but the content of projection may or may not be known to the person as part of himself; 4, projection can serve constructive as well as defensive purposes, can be normal as well as pathological. It plays a role in every act of empathic understanding and in such cases need not be distorting."(*)
I normally think of a proposition as an assertion with a subject and a predicate. I'm not sure I understand how an assertion about a subject can be "attributed" to others in the way implied by your substituion of the word for 'projection'.
We can mishear what another person says, and we can then embellish our subsequent reports to others (gossip), and I believe this can be done both consciously and unconsciously. Responses to both the source and the subject of such "gossip" can be as if the "gossip" were absolute fact.
Consciousness of abstracting is supposed to be the way around such difficulties, but consciousness of abstracting can be abused also.
I infer, based on the way I understood your uses of "semantic reaction" that you have succumed to a common naive misconception. Virtually every response we have is a "semantic reactions" with the possible excepton of "knee-jerk" type reflexes. Every "learned" association produces a "semantic reaction". We react to every stimulus in terms of what it means to us as a result of our total history of association. Often they are consistent with other people's reaction, but we also all have highly idiosyncratic associations.
See Semantic Reaction.
Watch the spelling, please, the word used was 'percept', not 'precept'.
Now we have both, so the posts need to be clarified, as the meanings are very diferent.
Sorry, David,
Those were both in reference to your posts.
Your first post used both precept and percept, and I felt confused later on when you gave a definition of precept.
Our percepts get altered by our precepts. That is another way of saying the Whorfian hypothesis, or at least a corrolary of it. It's taken for granted in general semantics circles.
I would most likely be significantly more verbose and pedantic in my formulations. Filtering is somewhat like abstracting, but it unfortunately entails the notion that the "same thing" goes from one level of abstracting to another - a notion not consisten with the map (output of the filter) is not the territory (input of the filter). "Fabricate" contains an element of adding or misconstruing, however it normally connotes consciously and intensionally, the combination of which is too limited.
Yes, I misspelled intentionally unintentionally.
I still don't like 'filtering' as filtering implies a hetrogenious mixture in which some is removed allowing the rest to pass through. Built into this structure is an idea inconsistent with general semantics, that of "identity" of something passing through the filter. When we abstract, characteristics at one level stimulate or induce different characteristics at the next level. Light becomes photo-chemical recations. Photo-chemical reactions become nerve impulses. Nerve impulses become synaptic neuro-transmitter chemical movement. Neuro-transmitter chemical movements become nuro-transmitter binding to receptor sites. Binding becomes potential changes. Potential changes become new and different nerve impulses, repeated many time. some nerve impulses become muscular contractions. some muscular contractions become rushing air. Some rushing air becomes modulated sound pulses. Do I need to go on? This is all abstracting, but virtually none of it is "filtering".
Because also, the map reflects the map maker, some nerve impulses become semantic reactions of the type we might call memory recall, so when we are abstracting at verbal levels we change the words to new ones. It is rare that we simply leave out some words. This happens with percepts as well. Our history causes us to respond to some inputs by replacing them with memory recall. Sometime consciously, sometimes unconsciously, and often erroneously, as in the case with my earlier intensionally versus intentionally.
We don't just "filter" we delete (filter), change, and add. We don't just "fabricate"; we add consciously (fabricate), we add unconsciously, we change, and we delete.
That is half the story. "Delete" and "add" only account for one level of abstraction.
Change is representable using two levels of abstraction. At the more abstract level, that which is before is identified with that which is after. At the lower level of abstraction that which is before is deleted and that which is after is added.
The river before is identified as the river after.
Riverstep 1 (before) ceases to be (deleted) and riverstep 2 (after) comes to be (added).
Please review my earlier post.
We cannot say that something changed unless we are at one level of abstraction identifying something through the continuation of time, even though there is a difference between the earlier and the later time.
For the precise structure of "change" read {Organizing Knowledge} up to the paragraph beginning These structural elements and relations comprise the minimal structure needed in order to formulate the notion of a map. 
The notion of undefined terms was imported from axiomatic geometry. No such formalization is possible in natural languages, as we all have semantic reactions to every term, and every term has a time-binding history with the possible exception of new technical terms given explicit formal definitions in structured languages such as is used in computers (non-natural).
Moreover, the totality of terms is definitionally circular, as terms are "defined" in using other terms.
A few terms may be "undefined" in terms of explicit verbal formulations, but they are not without associations, connotations, informal "meanings", etc., to anyone who uses them, and the vast majority of "arguments" as to how terms should be used derives from these semantic reactions of individuals who cannot completely articulate their own understanding in precise formulations.
In natural languages terms may be labeled as "undefined" but that just means that their "meaning" in conversations is ambiguous, abstract, unspecified, inarticulate, open to interpretation, etc.
To try to model the axiomatic structure of formal mathematical languages by the mathematically naive is proving to be a disaster, as the vast repetition of failure to come to any kind of clear definition or formulation or even standard of use for such terms in general semantics discussions is clearly evident.
We talk, talk, talk, past each other all the time.
"Undefined" is euphemistically understood by some as "without any meaning". That is wrong. We have the time-binding record, and I advocate using google searches with define:word, for example, to get a good sampling of not only dictionary definitions but many other definitions - all verbal. If google define:word comes up empty, then you might have a term which "is" undefined (on the web). But even with Jabberwocky the nonsense words evoke sematic reactions in the listener. General semantics may wish to abdicate the responsibility to provide formulational definitions for 'structure', 'order', and 'relation', but it has not provided an axiomatic structure into which to embed these terms that implicitly defines them it terms of each other in the manner that 'point', 'line', 'plane', etc., are implictly defined in terms of each others in absolute geometry.
Hilgartner's attempt to formalize general semantics language using such structure was a total failure. See A Non-Formal Non-Language.
Note that if you interdefine three terms, each in terms of the other two, you have a natural language parallel to a system of three linear equations in three unknowns. Such a system can have a precise solution. This is an example of what I mean by "implicitly" defined, an the general semantics terms 'structure', 'order', and 'relation' are defined in this manner, although very sloppily and imprecisely, and, I might add, without official agreement.
When we "talk" in repeated feedback, we end up at a point where we ASSUME that we are both using the terms we think we agree on without further examination - and without further formulation. That is NOT the same as undefined.
Yes. My little "talk" begins with The "normal" mode of communicating is to (1) assume that one is using terms consistent with "standard" (dictionary) definitions within the universe of discourse and (2) that the listener has a high probability of understanding the transmitted formulation in terms of these standard definitions.... Since we typically "identify" our formulations, as expressed, with our own objective level semantic reactions, most of us natually assume that, in the absence of feedback, others "understood" (at verbal and objective levels) what we meant (if not what we actually said). Consciousness of abstraction alerts us to be aware that the "identification" is "false to fact" and that the "assumption" may be wrong.
The "empirical fact" that we can only define circles with terms is neither "empirical" nor "fact". I know of no one who has actually performed the alegged action that we all take for granted -- that there are a finite number of words, a finite number of sentences, and that the words can only be used once in a definition, so they will be "used up" before all words are defined.
The mathematical truth is that sentences can be as complex as you want, consequently there are actually a countably infinite number of sentences that can be constructed, of ever increasing length, and that words, once defined, can be used in definition sentences more than once. Consequently, no matter how many sentences we have, we can always add one more sentence that is about a previous sentence. So, there are, in principle, an infinite number of sentences, and therefore, we cannot "come to the end" of words in constructing definitions.
Some of the greatest advances in Mathematics have come as a direct result of a listener NOT assuming that he understood at objective levels what was meant - precisely by the term 'point', 'line', and 'plane'.
It was logical analysis without the encumberment of the "known objective level understanding" of these terms that allowed the discovery and creation of non-Euclidean geometries. The "intuitive" "objective level" understanding of these terms differ for positive and negative curved spaces. And, if that is not enough, we also have descrete geometries, and absolute geometries, all of which have arisen precisely due to "unblocking" by discarding the "assumed "objective level" "understanding".
My little "talk" ends with ... the conversation then needs a cycle of repetitive feedback to bring the formulations back down to an "appropriate" level of abstraction where either understanding and either agreement or disagreement is achieved or one of the parties decides to terminate the process without achieving understanding. I did not note it, but when one thinks one has achieved understanding, it is because one reaches a level where one assumes one understands - that is, one assumes that one is in agreement with the others on the use of a term (at verbal and objective levels).
A percept is the resultant of perceiving. It is the representation of an external event that affected the senses and which - by perceptual processing - caused the activation of a certain category in the mind, i.e., the percept. The term percept is typically used in contrast to the terms distal stimulus (the external object) and proximal stimulus (the physical stimulation pattern on the senses, e.g., the pattern of light wavelength projected on the retina off the distal stimulus). percept.
If you wish to extend the "concept" of "percept" to internal senses then internal senses that can stimulate "percepts" are proprioception and interoception, but that extension would not be consistent with traditional time-bound usage.
The "phantom limb" phenomenon would be an example of a percept originating within the nervous system, a percept, in this case, that was "false to fact".
Nausea is an example of a percept that comes from the brains interpretation of the non-muscular internal sensors - a paradigm case example of interoception. Others include the need to vacate, hunger, etc.
Proprioception is specific to sensing the position and orientation of body parts.
These two are, to my way of thinking, and my prior usage, mutually exclusive senses, just as sight and hearing are mutually exclusive.
When I was using the distinction, I was, and have in the past, remained consistent with Sherrington original distinction:
"As originally defined interoception encompassed just visceral sensations but now the term is used to include the physiological condition of the entire body and the ability of visceral afferent information to reach awareness and affect behaviour, either directly or indirectly." (*)
I guess "hunger" would be a percept under the extension of its application from just external senses.
To Thomas's question....
I'm inclined to think of a percept as a brain response at a higher level of abstraction than simply the perhipheral nervous system. If you initiate a knee-jerk reflex by tapping the tendon on your knee, you experience a "strange" timing sensation, because the leg moves before you can consciously command it to move. The reflex itself takes place entirely within the peripheral nervous system, but the proprioceptive sense is fed back to higher brain centers, and through adaptive resonance and autoresponding, stimulates the same neurological circuits that are also used in commanding the leg to jerk consciously, but the timing of those circuits and the proprioceptive feedback circuits is reversed in the two cases. As a result the two acts "feel" strangely different. One feels "normal"; the other feels "unfamiliar" and somewhat "abnormal".
Without a whole lot of such extensional examples, I don't think we can "define" percept any better.
I'm inclined to think that both cases initiate one percept - that of kicking, but clearly one could argue that they are diferent percepts.
It depends on the level of abstraction of the speaker.
A percept is not exactly something that we can hold up for "show and tell".
I recall earlier talk about percepts being "incorrect" with respect to the sensory input. That a person with a missing limb experiences a neurological response that creates the percept of kicking or of an itching (missing) limb is what I meant by "false-to-fact" percept.
In my case I have a minor disability that results in "referred" pain due to cervical area pressure on a spinal nerve. The mechanism is that non-sensory-caused nerve stimulation is interpreted by the brain as having been sensory-caused. The brain has no mechanism for processing nerve impulses that are generated as a result of non-sensory stimulation to the nerve trunk line; it processes the impulse as if they were coming from the sensory organs attached to the ends of afferent nerves.
Any phantom limb "percept" has no basis in a physical limb, so I labeled the percept as "false to fact".
The fact that all hallucinations have a physical neuological mechanism does not mean that they have a distal or proximal stimulus. They are "false to fact" in that regard, as they have no external physical referent.
Nora wrote ... interoception includes communication between parts of the brain. This is not precisely accurate, as interoception involves internal sensory processes, and the brain has no sensory process internal to itself. When the brain process neurological inputs from sensory - exernal and internal to the body (but NOT internal to the brain) - it relates these to its totality of memory (as Nora noted by communication between different parts of the brain) and we experience semantic reactions. The phantom limb experience can originate in afferent nerve activity (generated without the sense organs), but it can also orginate in the brain regions that have recorded past experiences of the limb as well as, in some cases, within the brain areas known as the motor and sensory homunculi. The later two cases are not examples of interioception, they are examples of projection without afferent nerve inputs. For it to be a case of interioception, the stimulus must be primarily from activity in the afferent nerve processes entering the brain. Because of the random firing nature of nurons and the reticular activation system, as well as hormonal activity, these afferent nerve process are always somewhat active to varying degrees, but the information content varies from the evolved coherent and recognizable (by the brain) patterns to chaotic random activity. Cut off the sensory cells that are at the ends of afferent nerve processes, and there is still some low level chaotic random stimulation going into the brain through the afferent nerve process. But the brain tries to make sense of theses signals in the only way it can. "In this circumstance, the sensation ... will generally be felt in the territory that the nerve serves, even though the damage [is] elsewhere." (Wiki)
At these levels the brain network can be likened to a telescope with holes punched in the side. We see light, but we don't know where it's coming from, but we still try to see what we think may be out there. With the scope functioning, our neurological processing filters out the "noise". But put a cap on the scope and you have only the "noise" to process. In trying to make sense of what we do see, we imagine we see things that we might normally see through the telescope. In this case the phantom limb experience as a result of non-sensory activity in the afferent nerve processes. Some stimulation can also come from the severed nerve ends, making the afferent activity strong enough that the brain readily process limb position, though distorted.
This all applies to the level of abstraction at which we divide the system into the brain and it's efferent and afferent nerves.
If we want to get down to a lower level of abstraction, then we can follow the afferent (and efferent) nerve processe into the brain regions that they serve and connect to. For normal motor and sensory neurons these areas are known as the motor and sensory homunculi, and they appear in brain pictures as distorted maps of the body. It's possible that a stroke or blood-clot in the brain can put pressure on some of these internal afferent processes, resulting in injury caused stimulation originating within the affererent nerve process after they enter the brain, but these are not due to afferent sensory organs and still do not constitute interoception.
How come I'm so verbose and pedantic?
It is important for understanding our model clearly to be precise with our definitions and usage of terms. The model is, of course, "validated" (corroborated) by experiment and other empirical studies, which may find flaws to change. But lack of precision in the definition and use of the terms is like trying to build a house out of soaking wet noodles. So, use your noodle; keep the use of your terms clear and precise. We want our model to be firm enough that when we break something we can discover it quickly.
It is only after several more levels of internal brain processing that we experience the projections that we used to think were the things out there in "reality". Most of those uninitiated into the "secret teachings" of general semantics still do think they are actually seeing what is out there. And, for day-to-day practical purposes, so do we.
PS. If "noise" does not work for you, because I've used two different sensory modalities in my analogy, substitute "light pollution" for "noise". "Noise" in information circuits is signals that do not contain information - like static in radios or "snow" on the older television sets.
David,
The answer to Thomas is a clear "no".
Exeroceptions gets processed in the brain, which recalls memory in the production of the semantic reaction. This is NOT interoception. See my previous post.
The only thing that I can think might be a round-about example is the case of a hypnotized subject told (exteroception) that he was burned who actually raised a welt on his skin. But without the actual physical welt, it would still be only a semantic reaction.
The other way around, you feel pain, then you turn your head and look, and you then see a burn. Now that is a different case, but the two process are still not directly connected.
Exteroception and interoception are clearly defined and distinct sensory path processes, and one does not lead into the other. They are as distinct as sight and hearing, as distinct as smell and sight. When they are processed together, it is only after they are being processed within the brain - like when we see something we are touching. The integration is an internal brain function higher up in the processing cycle than the afferent neurological processes.
I usually understand "fact" (or more properly "statement of fact") to be a proposition about some referent, usually in the external world. See define:fact.
So any proposition, such as "afferent nerves carry information about the external world", or "damaged afferent nerves generate signals that the brain misinterprets as coming from the sense organs at the beginning of the nerve path". are both "facts" or "statements of fact". They are empirically testable.
At one level all "perception" is projection, because it "is" the neurological activity going on inside the brain after the external stimulation, even though we "perceive" it as happening "out there". The brain is an organ that "locates" its experiences elsewhere.
Our senses vary in the way we perceive them.
Vision is projected "way out there".
Hearing is projected from far to near.
Touch is projected from right here to, in special cases, a little out there (blind people "feel" at the end of the cane;), but the feelings are mostly located away from our heads.
Smell is experienced mostly right here, but some of us can sense direction and some distance. The sense of smell - chemical and our oldest sense, almost right at the brain, "feels" almost in our heads.
"Percepts" not supported by "facts" range from "real illusions" to virtual illusions, to "slight of hand", and more.
I highly recommend Jeff Hawkins book, On Intelligence, as it provides a robust model that explains a lot. Our brains are continually predicting what we ought to see and experience next based on our learned history of prior experience and memory. The vast majority of what we "see" is the constructed anticipated prediction based on the last time we were in the same situation.
A small experiment you can conduct.
Take a person to the doorway of their private room with their eyes closed. Ask them to open their eyes for a second or so, and then describe the scene that they saw.
Perform the experiment twice. But for the second time, have someone make some rearrangements of a number of ordinary things.
The one second glimpse reports things to be not where they actually are, but where they normally are or where they were last seen by a prior more extended look.
The reports made turn out to be what the person's brain predicted the person would see, and it has many more errors on the re-arranged items, errors that report the items to be where they were remembered to be rather than where they were moved to without the subject's knowledge.
We "see" much more what we expect to see based on our past experience and memory than what we might see with a longer and more careful study.
Plenty of research on the unreliability of eye-witness testamony will corroborate my claim - some published in Scientific American decades ago.
We "see" a projection of what we expect to see.
We "see" a projection of what we believe is there.
We "see" a projection of what we want to be there.
The shorter the time of looking, the less accurate the projection.
All these are properly called perception - whether accurate or not.
When we "see" something that is not there while we are looking at it, if you will pardon the oxymoron, it's called a hallucination.
From more "accurate" to less accurate.
Perception.
Misperception.
Phantom limb.
Hallucination.
All are experienced in the brain, but with less and less sensory contribution.
That URL begins with Simply put, a Semantic Reaction is when one responds to things the way they "should" be, rather than to the way they are.
This statement illustrates a naive misunderstanding of semantic reaction. "Should" is irrelevant. Moreover we cannot know how things "are", as we only have our abstractions, so it also illustrates another naive misunderstanding of the structural differential and abstraction.
A semantic reaction is a response of any kind, "accurate" or not, conscious or not, that is to what the stimulus (external or internal) means or signifies to the person. The "message received" is not always the "message sent", and the semantic reaction is the message received. If you "see" a red flashing light approaching from the rear while you are driving, you typically would not be "clueless" about it; you would respond, perhaps verbally, perhaps rapidly pulling over without even thinking of any words. It both cases your semantic reaction is in terms of the "meaning" which we can express as "emergency vehicles have the rightofway" or ((getting out of the way fast)) (non-verbal).
But when you stopped besided the road and a low flying hellicopter passed you buy, you realized that your initial abstractions were "in error". Such is the nature of abstracting. We respond with our best model to explain our experiences, but we are continually correcting that model (map) with updated data. The history of science is full of examples of theories to explain the world that have later been shown to be wrong. The "same" mechanism applies to our own individual abstracting from WIGO. Our brains are continually building and updating a dynamic model to account for our sensory inputs. But our virtual environment - symbols, words, etc. - is much easier to get wrong.
Some projections project inappropriate semantic reactions, many projections project approriate semantic reactions.
I don't know about "translate", but the current school of thought, still the current philosophy of science, is more or less a more concise abstraction from Karl Popper.
Maxwell on "refusal to change". 
The basic premise of general semantics is that if we stick to the methods of science in "all" aspects of our lives, we will achieve sanity (over "unsanity") in "all" aspects of our lives. As such it is a belief system.
Using Jeff Hawkins's On Intelligence we can see that our "belief system" can be thought of as hierarchically organized and highly dynamic, manifesting itself even as low as moment to moment predictions. The lower in the abstraction hieararchy, the more easily corrections are assimilated to update these levels in the belief system. The higher in the hierarchy, the more accumulated anomalous abstractions are necessary to induce a "paradigm shift" - high level change of belief.
Moreover, the more higher level beliefs are "isolated" from empirical abstractions, the more difficult it is for them to be changed.
Religion and other superstitious beliefs and behaviors illustrate the difficulty with which such beliefs get changed. See Maxwell again.
But, and this is very important, these belief systems push down the hierarchy from the top, guiding both our behavior and our predicitions, many of which are untestable (isolated).
General semantics:
Become aware of the abstraction process.
Learn to identify assumptions.
Test our assumptions.
Discard ones that cannot be tested.
Hold all beliefs conditionally.
In other words, join the "religion" of Science. 
I don't think that one can not have a perspective. That's the essence of relativity; we always have a perspective because there is no preferred frame of reference.
"By the use of reducto ad absurdum, all views can be proved false." is a view that supposedly can prove itself false, consequently it is inconsistent.
If there is only one choice, then there could not have been alternatives.