Of these two kinds of words, iḍukuṛippeyar and kāraṇappeyar, the first alone will serve the purpose of this enquiry. They alone come down from the far off ages when the Tamil language was born, when objects and actions were named unconsciously or semi-consciously.
Other words will not serve our purpose. Modern Tamil vocabulary includes words borrowed from English, French, Dutch, Portuguese, Persian, Arabic, Sanskrit and Prakrit. Of the loan-words from Sanskrit, some have been borrowed wantonly, i.e., when there are many Tamil words to express the ideas; this was partly due to Brahmanas whose familiarity with Sanskrit made them import such words in their Tamil speech and writing. This extensive borrowing was also due to the necessities of rhyme and assonance, a great characteristic of Tamil poetry. Loan-words began to enter Tamil not before 1000 b.c. and cannot be of any use in investigating the life of the Tamils before they came into contact with other nationalities, except that words not wantonly borrowed may be used as negative evidence to show what the Tamils were not acquainted with before such borrowing. But the date of these borrowings cannot be fixed. So even this negative evidence is not of much use. Similarly what are called kāraṇappeyar, words deliberately invented to name things and express ideas for which there were no iḍukuṛippeyar, cannot also serve our purpose, for such casual names can be invented at any stage of a language and cannot be proved to have existed or to have not existed at any particular period of time.
Hence iḍukuṛi names alone will be used in this enquiry. Such words in Tamil are practically root-words, without the wrappings of prefixes, augments, suffixes, etc., which disguise the root in Sanskrit words and make Sanskrit etymology so difficult and in some cases unconvincing. As these iḍukuṛi words are naked root-words they belong to the earliest stage of Tamil, the stage when the language was unconsciously forged by the stone-age man. Examples of such words are maṇ, pul, uṇ, pō, tī, nīr, mīn, vān, ā, kā, etc. The stage of the invention of such simple root-words cannot occur more than once in the history of a language. First because it is a stage of unconscious development of a language; secondly, if roots could be invented at any stage of a language, there would be no necessity for loan-words and consciously invented compounds at all. When men after progressing beyond the earliest stage of a language found or made new things which required names, the native power of inventing roots having become exhausted, they semi-consciously extended the meanings of old words by the processes of metaphor and metonomy. Examples of words which belong to this stage are, maṛai, shield, from maṛai, to hide, pon, metal from pol, to shine, śembu, a pot from śembu, copper, itself from śe, red. This may be treated as a second semi-conscious stage of the development of a language. These words are practically iḍukuṛi words, and will be utilized in this enquiry. A language becomes fully conscious only when it comes in contact with foreign languages; then it finds its soul, as it were, and becomes conscious of its structure; then alone it forges compound