Saturday, July 17, 2021

Thoughts on Roshomon

 My niece sent me the following TED clip about the Roshomon Effect and the apparent conclusion that we can't know anything for sure.  Talking in absolute terms I would agree (Roland's Flock) but outside of "absolute truths" I think that we are still able to determine true from false, fact from fiction and just because the evidence provided by the various observers is contradictory or possibly mistaken does not, at least for me, mean that all is lost and we have to swim in a perpetual sea of speculation where we can't know anything.

 


In the clip each observer is describing their experience of the unknown object in terms of known objects so, for instance, the observer that it touching the elephants side knows what the experience of touching a "wall" is and is drawing a comparison between that experience of a known object with the unknown object at hand. When using the word "wall" to communicate what her experience was she is relying on a shared knowledge of what a "wall" actually is and the assumption here is that what is meant by a "wall" is the same to everyone.  This is not necessarily the case but it can be resolved by getting all observers to experience for themselves what the one observer means when describing the experience of a "wall" is.  One could make the remark that this is the same as the conundrum of the elephant and leads to infinite regression but I don't think that this is the case because on the one hand one is describing the unknown (elephant) in terms of the known (wall) whereas on the other hand one is defining what is meant by the experience of touching a "wall". By having a shared experience of a "wall" between the observers each can understand the experience that the one observer had when describing the unknown object as a "wall" or a "spear".

Having a shared vocabulary of what each observer experienced is key because it enables us to use this to at least eliminate things that the unknown object can't be.  For instance, the unknown object can't be a "wall" or a "spear" because we know what a wall and a spear are, by shared common experience, and they are not the same so we can now make the move from the unknown object being either a "wall" or a "spear" to an understanding that it is neither and that what the two observers experienced was simultaneously "like" the experience of a "wall" and a "spear".

This now puts us in the position that we are searching for an object that fulfills all the observers experiences at the same time; a bit like a jigsaw puzzle with one piece missing and hundreds of candidates that may or may not fit and even if one does fit, it may still not be the correct piece because although some properties e.g. it fits in the physical space may be satisfied, the picture on the piece is inconsistent with the pieces surrounding it.

Going through possible options for the unknown object we would be able to eliminate many that don't fit any of the observers experience let alone all of them; for instance we can confidently eliminate "an ocean" or "a sun" and so on that don't have any of the properties that we are expecting the unknown object to have.

We are now left to work through objects that fit one or more of the observers experience but because they don't fit all of them, these too can be discarded as candidates. So for instance an option of a "rhino" may fit most but not all of the experiences the observers had so; close, but no cigar.

Having eliminated these as candidates we are left with objects which fulfill all of the experiences. In this example lets say, "elephant" fulfills all the requirements but is that an "african" or an "indian" elephant; we have no way of knowing from the observers experiences and more experiences would need to be taken into account to determine which species it happened to be.  So we would need to be satisfied with the provisional answer of "elephant" in the absence of any further observed experience.

Finally, the possibility exists that we can't find a known object that fulfills all the requirements. In this case we are left to ask if any of the observers were perhaps mistaken or, if that is not the case, have we simply incorrectly discarded a candidate in the process leading up to this point or are the experiences that of separate objects rather than a single object.  All of these are options that need to be assessed before entertaining the final possibility that this is an as yet undiscovered object having all of the properties experienced.  The previously unknown object now becomes known by common experience as something that has all of the properties experienced and even though we don't know all of the properties of the object we at least know some of them.  This illustrates, at least for me, the constant progression from the unknown to the known even while acknowledging that we don't, and can't, know everything about an object be it a "wall" or a "spear" or the unknown object at hand.

For me the Roshomon example illustrates how we should be wary of our conclusions rather than precluding us from arriving at conclusions at all.  In addition, by illustrating the provisional nature of our knowledge it also points to the repetitive nature of re-visiting previous conclusions as other evidence and experience becomes available.

It is for the above that I am of the opinion that we do not need to swim in a sea of speculatation and that knowledge can can stand on the firm ground of a growing continent even whilst acknowledging that that continent itself is in motion through a sea of speculation.