The Ultimate Dictionary

Atkins writes about how difficult having multiple word-forms with the same meaning makes lexicography. I find this interesting because I wonder if this means that some languages are easier to study than others. English, for example, has many synonyms whereas other languages only have one word for many things. It is interesting that the hyponyms to shake – quake, quiver, shiver, shudder, tremble, and vibrate – are often not used in dictionaries’ definitions of shake. This raises a question of how difficult it is to keep updated and write dictionaries. Atkins reminds us that dictionaries will never be perfect. However, what could be the most efficient way to write a dictionary? I would assume it would have to do with programming and using a constant flow of data from social media. This would have to use dictionaries to try to link hyponyms. Linking hyponyms in an online dictionary would allow users to more easily use a thesaurus and it would aid linguists in understanding language. In addition to linking hyponyms for linguistic purposes, linking word families would also help. Word-forms in word families are listed as different words in dictionaries. Complex lexemes are also listed separately but this should probably hold when collecting data. Different word-forms in word families can be linked in this possible dictionary maker because they denote a similar way of speaking and people who say one word-form are likely to say another. To make this dictionary program we would need to choose what rules to follow. Would hyponyms like the shake words be linked or would we consider these complex lexemes and not link them? Another rule we must consider is how to deal with compounds. Should these words be linked to the head or the dependent? Should they not be linked at all? They obviously mean different things but the data from the two separate members could be helpful for linguistic assessment. In “The Many Ways to Search for a Frog,” Slobin teaches us that diversity of language is not as boundless as we may think. Because this is true, this dictionary program is actually possible. Typological linguists could collaborate to find variables on which we can build the program. We would be able to get a diverse set of data from all over the internet including audio, video, and texts. With all this data, through use of various parameters, we could create a super dictionary that is most often correct.