# On rituals and algorithmsJuly 17, 2012 7:10 AM   Subscribe

What is the relationship between the ritual and the algorithm? Are all rituals algorithmic?

I'm interested in any papers or books/chapters written about this. I'm writing about Google's search algorithm at the moment, and I can't help get this ritual thing out of my head. It feels like an important meeting point between humans and machines.

When I say 'ritual' I mean the enactment of a set of actions with traditional and symbolic value.

When I say 'algorithm' I am talking about computers of course (a step-by-step procedure for calculation), but I am also interested in an algorithm as a list of well-defined instructions passed on to another entity in order to execute a specific procedure in precise detail. That entity might be a human.

Are all rituals algorithmic?

The best conflation I can think of is the Japanese Tea Ceremony. To generalise, I see this as a highly specific algorithm taken to the absolute limits of human cultural perfection.

posted by 0bvious to Religion & Philosophy (16 answers total) 10 users marked this as a favorite

Are all rituals algorithmic?

Not even vaguely. Many rituals are participatory and unstructured, or have loose boundaries rather than a step by step checklist of tasks to perform.

I might even argue that once a ritual becomes so codified and hidebound that it can be performed from a checklist, it's lost most of its ritual power. (exceptions may be cases where the act of following a checklist is the ritual -- the tea ceremony may fit into that but I'm not really that knowledgeable about Japanese culture.
posted by ook at 7:25 AM on July 17, 2012

The defining characteristic of an algorithm is that it accomplishes something. Once you're finished, you have created a result which you want. An algorithm is a means to achieve some end.

That isn't really true for rituals. It is the act of participating which is important, not the result when it's finished. I don't see the two as being related in any significant way.
posted by Chocolate Pickle at 7:29 AM on July 17, 2012 [4 favorites]

By trying to make a ritual into an algorithmic formula, you are eliminating the heart of the ritual itself.

The Japanese Tea Ceremony is much more than a "list of well-defined instructions." It is a meditative, spiritual process of being in harmony with your actions, being focused on the moment, and being mindful of your self.

It is like trying to find a formula for falling in love - it won't work. The murky inner working of the human mind are not so easy to turn into a formula.

If you try make a formula out of "being mindfully present in the moment" - you are missing the point.
posted by Flood at 7:41 AM on July 17, 2012 [1 favorite]

Most human rituals are not an algorithm, at least according to my working definition of what an algorithm is.

(1) An algorithm is a finite set of specific instructions. Religious or spiritual rituals may fail this part of the definition. For example, Waiting worship for the Quakers fails this test: they all arrive at church together (specific step 1), then wait for God's voice to reveal itself (nonspecific step 2). Apologies if I misunderstood the actual ritual. However, other religious practices (the act of taking communion for example) have well defined steps. Psychological rituals such as many OCD rituals have specific steps as well, as do the social rituals of the tea ceremony.

(2) An algorithm produces a well-defined output. In the sciences, this output usually takes the form of a mathematical expression, but that's not always the case. Religious rituals may or may not satisfy this part of the definition, depending on your perspective. The process of taking communion is well defined, but what is the end result? God's grace? Eating bread? Some more personal or individual result? Well defined end results are (I think) apparent for OCD rituals (clean hands, for example) and tea ceremonies (a cup of tea), but may not be for religious rituals.

Some rituals do satisfy all of the definition of the word, which brings me to the "working definition" part of my answer:

(3) An algorithm is intended to be efficient. The goal of every algorithm I've ever used is to acquire the desired end result quickly. For this reason, I would say the OCD rituals or tea ceremonies are not algorithms, at least in the sense of what one sees in the sciences. They are designed specifically to be inefficient, in order to satisfy either a psychological need (unnecessary washings to relieve anxiety) or cultural need (to be honest, I'm not clear on why the tea ceremony is performed). The equivalent for google would be if they required additional, unnecessary steps (asking your favorite color, or requiring you to turn on and off the lights for example) in order to produce the search results you requested.

I think rituals that do satisfy parts 1 and 2 of the definition, but not 3, are interesting in their own rights, because a desire for inefficiency says a lot about who we are (sometimes beautiful, sometimes strange, and sometimes both). I wouldn't compare them to algorithms.
posted by bessel functions seem unnecessarily complicated at 8:10 AM on July 17, 2012 [2 favorites]

I expect that all human rituals are embarrassingly algorithmic. I'm reminded of the digger wasp; and have to wonder the ways in which this question bleeds into the realm of muscle memory and procedural memory. It strikes me that the question is complicated especially by the existence of higher level decision making processes in the human mind outside of the ritual, and because of this it would need to be approached stochastically.
posted by Algebra at 8:44 AM on July 17, 2012 [1 favorite]

ook: I might even argue that once a ritual becomes so codified and hidebound that it can be performed from a checklist, it's lost most of its ritual power. (exceptions may be cases where the act of following a checklist is the ritual -- the tea ceremony may fit into that but I'm not really that knowledgeable about Japanese culture.
You're discounting the ritual power of rosaries, masses, traditional wedding ceremonies, traditional funeral services, bar mitzvahs, bat mitzvahs, brisses, officer promotions, commendation and award ceremonies, the singing of the anthem prior to sporting events in the U.S, the swearing-in of government leaders and new citizens... Most of the world, in fact, seems to disagree strongly with you.
posted by IAmBroom at 9:00 AM on July 17, 2012 [2 favorites]

I think rituals that do satisfy parts 1 and 2 of the definition, but not 3, are interesting in their own rights, because a desire for inefficiency says a lot about who we are (sometimes beautiful, sometimes strange, and sometimes both). I wouldn't compare them to algorithms.

But sometimes we create mathematical algorithms that are inefficient just because we find them strange or interesting. For example, bogosort is a method of sorting a list of numbers based on randomly shuffling the list until you get a list that is, by coincidence, sorted. Or then there's spaghetti sort, a sorting method using pieces of uncooked spaghetti.

There are clearly bits of human behavior that we are concerned with optimizing (e.g. assembly line labor, washing the dishes). And then there are parts that we aren't (e.g. religious rituals, Rube Goldberg devices). The same is true of algorithms. Sometimes we are about efficiency, sometimes it's okay just to have an interesting result.
posted by jedicus at 9:06 AM on July 17, 2012

bessel functions seem unnecessarily complicated: (1) An algorithm is a finite set of specific instructions. Religious or spiritual rituals may fail this part of the definition. For example, Waiting worship for the Quakers fails this test: they all arrive at church together (specific step 1), then wait for God's voice to reveal itself (nonspecific step 2). Apologies if I misunderstood the actual ritual. However, other religious practices (the act of taking communion for example) have well defined steps. Psychological rituals such as many OCD rituals have specific steps as well, as do the social rituals of the tea ceremony.
Sub DoQuakerChurchWorsip {
1. Goto Church
2. WAIT_TAG:
3. If Not GodsVoiceIsRevealed Then Goto WAIT_TAG
}

Seems pretty well-defined to me.
(2) An algorithm produces a well-defined output. In the sciences, this output usually takes the form of a mathematical expression, but that's not always the case. Religious rituals may or may not satisfy this part of the definition, depending on your perspective. The process of taking communion is well defined, but what is the end result? God's grace? Eating bread? Some more personal or individual result? Well defined end results are (I think) apparent for OCD rituals (clean hands, for example) and tea ceremonies (a cup of tea), but may not be for religious rituals.
You actually provided your own rebuttal here, curiously.
(3) An algorithm is intended to be efficient. The goal of every algorithm I've ever used is to acquire the desired end result quickly. For this reason, I would say the OCD rituals or tea ceremonies are not algorithms, at least in the sense of what one sees in the sciences. They are designed specifically to be inefficient, in order to satisfy either a psychological need (unnecessary washings to relieve anxiety) or cultural need (to be honest, I'm not clear on why the tea ceremony is performed). The equivalent for google would be if they required additional, unnecessary steps (asking your favorite color, or requiring you to turn on and off the lights for example) in order to produce the search results you requested.
The goal of an algorithm is not to acquire the desired end result quickly. Rather, the goal is to apply a consistent methodology or process, regardless of efficiency. There are sorting algorithms that are far less efficient for some sorting problems (for instance, resorting "zyxwvut...cba" into alphabetic order again); they don't take the most efficient path, but the path that they are coded for.
posted by IAmBroom at 9:09 AM on July 17, 2012

Since people seem to be inventing their own personal definitions, it might be helpful to establish what the word actually means. Here's one actual definition:
al·go·rithm noun
a set of rules for solving a problem in a finite number of steps, as for finding the greatest common divisor.
This does describe many, but not all, human rituals, to varying degrees. The consecration of the host in the Catholic Church is so invariant that the priest is required to read each word individually from the liturgy, to avoid memory errors from creeping in. In contrast, many modern wedding ceremonies are decidely not repetitious (even from ceremony to ceremony).
posted by IAmBroom at 9:19 AM on July 17, 2012

Penn and Teller may think otherwise, but the 12 steps of AA are as close to an algorithm in a ritualistic form as may be had. People may vary the steps and interpret them differently, but when they are followed thoroughly, abstinence should result.
posted by Xurando at 9:30 AM on July 17, 2012 [2 favorites]

Sorry, I was unclear. The behavior after god's voice was the unspecified part, not the waiting part. It would be more along the lines of

sub DoQuakerChurchWorsip{
GoTo Church
while(!godsVoice){
wait
}
Interpret_Gods_Voice
Do_Something_In_Response_To_That_Interpretation
}

I'd say the latter two are decidedly non-algorithmic. You are correct, though, going to church and waiting are clearly steps. Sorry for the poor wording. You've also highlighted God's grace as the output apparently. I'm not clear why this is the output of the ritual (which is why I listed others), so I don't think it's obvious it has a well-defined output.
posted by bessel functions seem unnecessarily complicated at 9:34 AM on July 17, 2012 [1 favorite]

A possible connection: "It has been suggested that the art of counting arose in connection with primitive religious ritual and that the ordinal aspect preceded the quantitative aspect." A History of Mathematics by Carl B. Boyer, Uta C. Merzbach, Chapter 1.

The general idea is that way back when, some primitive culture had a ritual that re-enacted (say) the creation of the world. They would have some formulaic words that introduced each section of the ritual, which always came in the same order, and cued participants to do various things.

Since these always came in the same order and everyone knew them, some of these words (maybe the initial words of each section of the ritual) came to mean first, second, third, fourth, and so on.

FWIW the origin of the musical solfege syllables "do re mi fa sol la ti" is similar--the original "ut re mi fa sol la" were first syllables of musical phrases of a hymn where each phrase started on the corresponding note of the scale.
posted by flug at 9:39 AM on July 17, 2012 [3 favorites]

You're discounting the ritual power of rosaries, masses, traditional wedding ceremonies, traditional funeral services, bar mitzvahs, bat mitzvahs, brisses, officer promotions, commendation and award ceremonies, the singing of the anthem prior to sporting events in the U.S, the swearing-in of government leaders and new citizens...

Not at all. Very few of those are by-the-numbers checklists: nearly all of them have a tremendous variety in how they're performed (even within a given culture or religion, in most cases.) People write their own wedding vows, funerals are pretty unique, public ceremonies and swearings-in are tailored to the person and the situation... they don't look to me like do-this-then-do-that algorithms.
posted by ook at 2:01 PM on July 17, 2012

People write their own wedding vows, funerals are pretty unique, public ceremonies and swearings-in are tailored to the person and the situation

By 'traditional' I think the commenter was referring to mainstream Protestant and Catholic services, which are pretty specific (e.g. the Methodist marriage service [pdf]).

Swearings-in are tailored but still scripted. For example, the presidential oath of office is precisely specified in the Constitution. Swearing in a witness uses a different script, but it's still a script.
posted by jedicus at 2:23 PM on July 17, 2012 [1 favorite]

[Folks, minutiae of different religions is well past on-topic for this thread, please take side discussions to email]
posted by jessamyn at 6:05 PM on July 17, 2012

Thanks for the comments. I'm still not convinced that rituals are not (usually) algorithmic at their core. The distinctions people have raised here, about the ritual embodying some meaning beyond the mere act, still stand true. For instance, a robot might deliver the most gesturally perfect tea ceremony ever, but it wouldn't be involved in any meaning arising from it. But this might be where my interest really sets in: the ability to lose oneself in some kind of meditative state induced by the ritual is, for me, a pure reduction of the human to the algorithm. The messy human-ness of things gets the way. Perhaps rituals bind us in the algorithm, in the pure flow of things.

I don't think rituals deliver mindfulness, I think they aim to deliver mindlessness. That sounds pretty algorithmic to me. I think there's something here about auto and allo-poiesis, but I'll save that pondering notion for another AskMefi question.

Thanks again!
posted by 0bvious at 5:14 AM on July 18, 2012

« Older Looking for South African novel   |   Can or should I go back to college? Newer »