Lexical entries in the Glossary

English spellings

The first part of the Lexicon, labelled English, lists forms of the name or word that are used in an Anglicised transliteration, i.e. with nothing more than the letters of the English alphabet. Note that occasionally one or more letters with diacritical marks (mostly a macron, a small horizontal line over a vowel) may appear in the glossary headword and among the English forms. This only happens when the fully Anglicised forms of two distinct names would look the same; in such cases distinction is made possible by adding the diacritics. For example the god Krishna (Sanskrit kṛṣṇa, masculine) and the river Krishnā (Sanskrit kṛṣṇā, feminine) could not be told apart if both were simply Krishna in English. This is not necessary, nor even actively encouraged, in non-technical texts, but is necessary for the site to work properly and to tell you correct information.

The recommended Anglicised form (in almost all cases identical to the glossary headword) is the simplest and most straightforward English transliteration of the accurate original spelling. This is not always what is most commonly used by the rest of the world, but is the form that appears consistently on this site and should in my opinion be used in all English-only environments. Very rarely, several recommended forms may appear; these will be explained under Essentials in the glossary.

For some glossary entries you will also see an acceptable English form. The main reasons why different spellings of the same word may be acceptable are:

• the same word may be used in different grammatical forms (e.g. Sanskrit nominative case rather than Sanskrit stem as usual);
• a spelling based on the pronunciation of the word in a language other than the language of the article (e.g. Hindi or Tamil) is commonly used by English speakers; or
• the original pronunciation of the world can be equally well approached by several Anglicisations.
  

For most glossary entries there is also a discouraged section that lists English spellings of the word that are or have been in use but should not be. Most of these are dated Colonial transliterations, while others are simply non-standard or plain wrong.

The last English item under most headings is one called variations. This lists common Anglicised transliterations of the pronunciation of the word in some present-day Indian languages. Most of the Glossary items are in Sanskrit, the ancient language of Indian culture. Many modern Indian languages are related to Sanskrit, while most others use lots of words borrowed from that language - but most of these have their own pronunciation rules that are different from classical Sanskrit. Often, especially when Indians use Sanskrit words, one of these modern-day pronunciations will creep into English text. This is not a good idea when dealing with words steeped in tradition, and if you are a westerner, I'd discourage you from using these forms, but you might encounter one of these spellings on the net or in books. For a parallel, think of Latin: Latin names and words are pronounced differently in Italy, England and Russia, but when you talk about ancient Roman gods, cities or whatever, you would want to use the ancient pronunciation. Italians will say these things the Italian way, but even though Italian is now spoken in Rome, the rest of us should not follow their example.

If you're interested in where such a variation might come from, there are a couple of typical changes that betray this. For example if a variation lacks a short a that the Sanskrit version has (usually at the end of the word, sometimes in the middle too), then that variation is probably from Hindi which, along with most modern North-Indian languages, does not pronounce the short vowel a in an unstressed position. If the variation involves a change between voiced and unvoiced stops formed at the same place (e.g. t/d, k/g) or if it changes t or d to th or dh, then it almost certainly comes from Tamil, a language where voiced and unvoiced stops are not distinguished in writing and where there are no aspirate stops (th, kh, ph, etc.) so consonant+h combinations in English script are used for a different purpose. Many variations will also involve a simplification of consonant clusters (multiple consonants next to each other), either by fusion into just one consonant or by insertion of vowels inside the cluster. This is done in one way or another by practically all currently spoken languages. Finally, you'll also see that for some words the Sanskrit nominative case is different from the Sanskrit stem (more on this below), and in most of these cases the modern Indian variation will be closer to the nominative than to the stem form.

An interesting side effect of having the modern variations of words in the Lexicon is that you'll be able to understand many Indian personal names if you look at these. 

Original language

Special characters

The Original language part of the Glossary Lexicon uses a couple of special characters. Transliteration appears in a sans serif font to set it apart from English text; Devanagari and other Indic scripts will appear in whatever font on your system supports these characters. Both transliteration and Indic scripts use the Unicode UTF8 encoding scheme. If you see garbled symbols in place of parts of these entries, try to tell your browser to reload the page using UTF8 (Look for View and Encoding in menus). If you get hollow squares instead of some of the letters, then your browser has got the encoding right, but you don't have a font installed that supports the required glyphs. Try to download and install Tahoma, Arial Unicode MS or (on a Mac) Lucida Sans Unicode. For Indic scripts, install any font that supports that script in UTF8. More help with fonts will be available on this site later on.

Transliteration

The transliteration block tells you the exact spelling of the word in its original language, using an internationally established scheme of diacritical marks for the transliteration of Sanskrit, with minor deviations for other languages. (More info on this will be published on this site later on; till then, see Wikipedia.)

Sometimes you'll see a second transliteration in square brackets. This tells you the Sanskrit nominative of the word, if different from its stem. In most languages the nominative case (the grammatical case used for the subject of a sentence, i.e. "Jack is coming") has no marker (or more properly, has a null marker), and is identical to the dictionary form of the word. Most words in Sanskrit do have a marker in the nominative (and the case that usually has a null marker is the vocative: the one used in addressing someone or something, i.e. "Hey Jack, come here!"). When dealing with Sanskrit words in a western-language context, it is standard practice to use the stem or dictionary form of words in the neuter or masculine gender. For most words (those that have stems ending in a vowel) the case marker for nominative is either nothing, or -ḥ in the masculine and -m in the nominative. This is usually disregarded in transliteration. The feminine nominatives of such words usually involve a lengthening of the vowel at the end of the stem, and if the word is a feminine noun, then this ending is retained in Latin transliteration. Some words (with particular types of stems) involve a rather stranger change in the ending. More on this will be available on this site later on. For the glossary, all you really need to know is that when you see a transliteration in square brackets, it tells you the form of the Sanskrit nominative. North Indian languages usually use Sanskrit words of the last category (strange change in the ending) in the nominative form, so when words are borrowed from these into English, this form will often be reflected in writing - though it should not be if we are dealing with proper Sanskrit (see my remark on Latin above). On the other hand, South Indian languages will often keep - and transcribe into English - the -m ending of Sanskrit words in the neuter, which again, should not really be done.

Analysis

The Sanskrit language has a particular fondness for compound words and complex derivations. A notable feature of Sanskrit spelling comes to the fore when dealing with this: the end of a word is often merged with the beginning of the next word to reflect natural pronunciation in writing. This is called sandhi, literally 'joining', and is also done for separate (not only compounded) words in the same sentence. Essentially it's as if in English you wrote "Jakkomes" instead of "Jack comes" - that's what you say, and in Sanskrit what you write is what you say. (There's also real sandhi in English too, although it isn't called that: for example "he's" for "he is" is absolutely proper sandhi, and writing "the idear is" for "the idea is" or "a haristocrat" instead of "an aristocrat" is the not-quite-standard reflection of sandhi phenomena in certain pronunciations.)

But back to the Glossary. When the heading is a compound or suffix-derived word, a separate block under Original Language will tell you what elements it is composed of, also resolving sandhi on the way, so you'll be able to see the elements in their unjoined, pre-sandhi state. Three separators are used to set apart three levels of closeness of the combined units; this is particularly useful in long compounds of multiple elements.

: (colon) – separates the distinct but close-knit elements of smaller compounds inside a larger compound.

. (dot or period) – separates suffixes, prefixes, or other elements (fixed morphemes) that are not meaningful words on their own, but change the meaning of whatever they are attached to.

Indic script

Words of Indian languages will also be shown in a native script. For Sanskrit (as well as most modern Northern languages) this script will be Devanagari, though strictly speaking, Devanagari is not "the Sanskrit script" as is believed by many. In modern India Devanagari is most often used to write Sanskrit, but over its history of some three thousand years and a million square kilometres, this language has been written in lots of different scripts, and any of these could be tagged "Sanskrit script". Unlike most languages we know, there is no single script attached to Sanskrit. Tamil words will be shown in Tamil script. In the future I might introduce Arabic script for Urdu words but for now, Urdu will also be represented in Devanagari, as a sort of generic Hindustani. Other Indian scripts such as Bengali may also appear later on in some of the glossaries.

As for Transliteration above, the Indian script version also uses the stem of words if that is distinct from the nominative. This is not standard practice in India (where the nominative will be preferred), but suits the needs of the Western Informed Outsider better because it is consistent with our tradition and with dictionaries.

Literal meaning

The last Lexicon entry in a Glossary will give you a short literal translation of the word. Most Sanskrit words have gathered meanings like moss over the ages, while some - particularly those pregnant with religious or philosophical meaning - just don't translate well into English. To complicate the matter further, Indian intellectuals over the ages have enjoyed a sort of word-play that analyses words into tiny parts that have nothing to do with the actual origin of the word, but shed light on or embellish its sense as used in a particular context. This is called popular etymology, and it is indeed popular in the sense that many people today will believe some of these derivations are the actual origins of Sanskrit words. To show you what I'm talking about, an Indian "popular etymology" for the English word god says that it's actually an acronym of the primary functions of the godhead: Generate-Order-Destroy.

All in all, giving a single literary translation of such words is a tacky business. I always try to choose the shortest and most down-to-earth meaning for this entry, but other meanings are almost always possible and in some cases actually better. So if you see an Indic word interpreted another way elsewhere, that does not necessarily mean that either my Glossary or your other source is wrong.