Sunday, November 15, 2009

The Number Seven, WordNet, and Assuming that Others are Being Honest

I ALWAYS HAVE a number of dictionaries nearby when I do any significant writing. Near at hand is a compact Ecclesiastical Latin dictionary (which includes all words used in the Vulgate and liturgy of the Church), an excellent French-English dictionary, a priceless etymological dictionary (which goes deeply into the word-origins of modern English), and a geographical dictionary which has translations of geographical terms in many languages. Out of sheer convenience (or perhaps laziness) I most often use the dictionary and thesaurus that comes with my Mac computer, despite the fact that its choice of example sentences is often annoying due to its political correctness. I also use various online Internet dictionaries.

While using the right word for the right situation is crucial for good writing and for making oneself clear, I don't claim that my own writing is particularly good or clear, but sometimes I try.

Lately I've been attempting to come up with a good name for a project. A good name is of great importance, as I've written about in the articles O, be some other name!, and Naming Places. In this pursuit of finding a good name, I've asked for advice and have done Internet searches on the principles of naming. Common advice is that a name ought to be short, unique, memorable, meaningful, and applicable. My problem is with short names, since there are lots of good long names.

So how long is too long? I found an excellent article, The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information, by George A. Miller, published in 1956. This is reported to be one of the most-cited papers in psychology, and it states that there seems to be a built-in limit to our judgement and and short-term memory, being roughly seven categories or items, plus or minus two (although Miller and his critics all state that we shouldn't use this result uncritically).

But I digress. Finding Miller's paper led me to the other things he has been doing more recently. Miller's current work includes most prominently WordNet, a lexical database, which uses relationships between words to develop what is simultaneously a dictionary and thesaurus, as well as a whole lot more.

WordNet acts both as a dictionary, giving definitions, and as a thesaurus, giving synonyms. But it also implements other relationships between words. Hyponyms are specific instances of a broader class, such as ‘dog’ is a hyponym of ‘animal’, while hypernyms are the classes to which a noun belongs, as ‘tree’ is a hypernym of ‘oak’. There are more relationships than these, including entailment, meronym, and pertainym relationships, which make WordNet a very powerful dictionary, difficult to concisely implement in printed form.

So how far do these relationships go? For example, I input the word ‘book’, and directed WordNet to give me all inherited hypernyms. For book (using the definition “a written work or composition that has been published (printed on pages bound together“)) WordNet states that a book is a publication; which is a work or piece of work; which is a product or production; which is a creation; which is an artifact; which is a whole or unit; which is an object or physical object; which is a physical entity; which is an entity. An entity is defined as “that which is perceived or known or inferred to have its own distinct existence (living or nonliving)“.

According to ancient philosophy, there are ultimately entities that are above physical and psychological categories, as the logic of WordNet shows. They have being in the loftiest sense of the word. The philosophical study of being is called metaphysics.  Now the word metaphysics has a bad New Age connotation these days, but that is only because scholars, particularly in the Catholic universities, largely abandoned that field of study back in the 1960s. As nature abhors a vacuum, others with less orthodox viewpoints made that term their own.

According to metaphysics, being transcends place and time, and has a universal unchanging quality. This does not suit the modern mind, and new philosophies are quickly developed, each new one claiming to finally jettison the problematic idea of metaphysics. Ironically, the old theories are usually abandoned because they have been shown to have a metaphysical basis. Hegel (died 1831) was called the last metaphysical philosopher, then Schopenhauer (d. 1860), then Nietzsche (d. 1900), then Heidegger (d. 1976); newer philosophers make plenty of metaphysical assumptions, and so they will be soon abandoned also.

One popular theory states that all these entities only exist in the mind, that they have only subjective and not objective being. But then what about when minds interact? What can we say about the similarity between the entities in the various minds? How and why are they similar? And so couldn't we say that these entities have something about them that transcends minds? Another theory states that nothing exists except atoms and impersonal forces.  So atoms have being? In a way that transcends space and time? After all, we can prove through experiment that an atom here is absolutely identical to an atom there. So why shouldn't other things have transcendent being too?

The ultimate in bad philosophy states that words have no meaning whatsoever, that nothing has any ultimate reality, and that there is no way we can know what is real or not. In this case, dictionaries are useless.  Why even bother? But many who hold this philosophy also claim that all words are just propaganda, used to enforce or change power relationships. But what is power? How do you know that your definition is the right one? What is the nature of a power relationship? How can you be certain that one person has power over another? Are there one or many types of power relationships? Why do you prefer one power relationship over another? How can you trust your own preferences?

Ultimately metaphysics is inescapable. I ought to note that people will often choose philosophies because it helps support their morality: we want to sin and so attempt to intellectually justify it. Instead of changing ourselves, we childishly attempt to change reality.

Debate can be frustrating, and we are often tempted to harshly judge our intellectual rivals.  George Miller has a wise piece of advice, known as Miller's Law:
In order to understand what another person is saying, you must assume it is true and try to imagine what it could be true of.
Calling someone a liar is a mortal sin, that being an offense against both justice and charity; according to Saint Thomas Aquinas, the sin of detraction is in degree just below that of murder and adultery. Likewise, a person's good name is the most valuable property they own, and in justice they may defend it to a proportionally great degree; these moral precepts are true insofar as the person is not a great public sinner.

If someone claims something apparently absurd, they are often judged insane or foolish, and not worthy of any more consideration.  However, they do deserve our respect.  I've noticed that a few probing questions — and careful listening — will quickly resolve or at least clarify the initial apparent absurdity, and you may end up having a fascinating conversation, as well as make a new friend.

The Medieval Schoolmen are often detracted for having excessive credulity.  But this is not true. Rather, like Dr. Miller, they assumed (out of justice and charity) that others — be they pagans of antiquity, Jews, or Muslims —  are basically honest.  The Schoolmen worked with heroic degree to parse out what is true, and to harmonize what others said or wrote, and this led to the great flowering of the Medieval Synthesis.

No comments:

Post a Comment