Instructions for Contributors — Russian Dialects



Guidance Notes for Making the Recordings

First of all, many, many thanks for your cooperation and enthusiasm in agreeing to help me with these recordings — it is much appreciated!  Once you send me your recording I will transcribe the words and add them to my internet database at  If you wish I will send you also a copy of the transcriptions and webpage files.

These notes should be all that you need in order to make the recordings for me.  I suggest you print them out to take with you for when you do the recording.


How Best to Organise the Recording Session

I have usually found that it is best, before making any recording, to go through the word list first with your informant.  This way you can both make sure in advance that your informant knows which words you are interested in, and that he or she ‘remembers’ the local pronunciation, as distinct from the standard language.  This preparation should not take more than about 20 minutes (depending on how much your informant wants to chat!).  If you do this first, the recording itself should then be very quick:  only about 5 minutes or so. 


Finding a Good Informant, & When to Make the Recording

The best approach is not to make these recordings at the beginning but towards the end of your fieldwork trip, when you have been able to identify the best informant for this task.  Also, by then your informant will be well aware that you are interested specifically in the local, and not the standardised pronunciation, and will have become more comfortable with talking in his or her local speech.  But please don’t forget!  :)


Practicalities:  How to Make a Good Recording

I normally use mono (not stereo) recordings, directly onto a computer (or .mp3 recorder) if possible.  I recommend the free recording and sound editing software Audacity, which can be downloaded free from;  you’ll also need this LAME converter file downloaded from  Alternatively, Windows includes its own basic recorder, to access it (on XP) do:  Start — All Programs — Accessories — Entertainment — Sound Recorder. 

Usually the most troublesome thing with computers is to get the microphone working.  Try it in advance:  the most common problem is that the microphone is muted (set to silent mode), so to check and fix this do:  Start — Control Panel — Sounds and Audio Devices — Volume — Device Volume:  Advanced — and make sure that the Microphone channel is not muted (silenced) by unselecting the appropriate box.

Any common sound file format will be fine, though .mp3 or .wav are probably best (I will convert them to .mp3 in any case, so that is simplest).  I normally record at 48KHz, usually the default setting for sound recordings.  A 5‑minute recording of the word list usually makes for a file size of about 4 or 5 MB in .mp3 format. 

In any case, for recording quality there are two things that are far more important than any of these technical settings:

   The quality of the microphone:  it doesn’t have to be an expensive one, but please do use a separate plug‑in microphone (the usual 3.5 mm audio plug) if possible, rather than just a ‘built‑in’ microphone hole in an .mp3 player or a laptop (the recording quality of those is very poor).  Microphone headsets (i.e. worn on the head with earphones) are also good, if your informant doesn’t mind wearing one.

   The level of background noise:  most plug‑in microphones are quite good at reducing background noise, but still please really try your best to find an environment where it is at a minimum.  If possible, be alone with your informant, in a closed room, in a house where not much else is going on.  A single sharp noise will ruin the recording of the pronunciation of a word, and it’s impossible to repair the sound file afterwards.  If you hear a sharp noise — footsteps, a door creaking or closing, a car passing, whatever — while your informant is reading the list, please ask him or her to repeat any of the words that might have been affected by the noise.


How to Elicit the Words in the List

In principle, it would be best not to prompt informants at all, neither by the list of the words in the standard written language, nor by your own pronunciation of them in the standard.  Images, for example, might be better, although for many of the words this is not particularly practical:  abstract nouns like winter, or specific morphological forms of verbs and adjectives.

Indeed in practice it takes up much more time to try to do this, and it is actually not really necessary.  What matters much more is to find a good informant — that is, somebody who understands clearly that you do not want to hear his or her ‘standardised’ speech, but his authentic accent/dialect.  Provided you find a good informant like this, who is able to make a clear distinction between standardised and ‘authentic’ local speech, then there is usually little problem of interference from the standard language on a good informant’s pronunciation.  Once he or she really understands that this is what you want, a good informant can normally simply work from the reference list, even if it’s in the standard written language.

The order of the words in the reference list has been specifically designed to help in this, starting off with the easiest words, and using symbols not written words:  the numbers 1 to 10.  These usually help informants immediately to get into the swing of pronouncing in the local way without reference to the standard.


Phonetic Matches Please, Not Meaning Matches

I use my recordings and transcriptions to measure the amount of divergence in phonetics between the cognate words in related languages — i.e. the net difference between their pronunciations that has arisen since the time of their common ancestor language.  I do not measure any differences in meaning.  My comparison is very strictly on the level of phonetics, not meaning. 

So, the word I need to have a recording of is the true cognate of the word in the reference list, not its meaning equivalent.  Please try to get the informant to pronounce his or her corresponding pronunciation of that word, even if it now means something rather different in his or her speech. 

In the Germanic languages, for example, it would make no sense for me to try to measure the phonetic divergence between German Hund and its meaning match in English, dog, because these words are not cognate — they do not go back to a common ancestor form (dog is a loanword into English from Scandinavian).  So instead, I measure the phonetic divergence between German Hund and English hound, the true cognates.  It is true that English hound has changed its meaning slightly, and now means specifically a hunting dog;  but for my comparison only in phonetics, this meaning change is not important.  So please record cognate forms, not meaning equivalents.  Thanks.


Missing Cognates or Influence from the Standard

The particular meanings and grammatical forms in the reference list have been chosen very specifically because true cognates of all of them exist still at least in all the official Slav languages, as far as I have been able to work out.  Nonetheless, it may be that in the dialect you are recording for me, the cognates of some of the words in the reference list are no longer used.  This might be the case with the word dog, for example:  perhaps in your region the root пёс is simply no longer used, and only cognates of собака are now known. 

If your informant gives a word that is not cognate with the one in the reference list, please do try to elicit the true cognate.  Nonetheless, if it is clear that it does not exist in that dialect, then just make that clear — I suggest you take a printout of the reference list and mark this in writing alongside any words that you are ‘suspicious’ of in this respect.  Please also indicate if you have reason to suspect that a pronunciation given by your informant is not a true dialectal form of that cognate, but reintroduced from the standard language.  This is particularly possible with the rarer or more ‘old‑fashioned’ words in the list. 


Morphological Forms

   My study is not a morphological one, but a phonetic one.  The word list is therefore intended to provide a balanced sample of the phonetics of the language, so it includes a range of grammatical forms so as not to repeat too many times certain particular sounds that occur in common grammatical endings (like feminine ‑a).  I have therefore varied the morphological forms of the words in order to create a balanced sample of masculine, feminine and neuter nouns and adjectives, in singular and plural forms.  (All have to be in the nominative case because the other cases do not exist in Bulgarian and Macedonian.)  Likewise with the verbs, there is a range of person and tense endings.  On verbs, I have tried to avoid extra perfective and imperfective suffixes wherever possible, because these often differ between the various Slav languages.  Please, therefore, try to record the exact cognate form of each word in the same morphological form as in the reference list.

   In most cases the exact morphological form I need is clear from how the word appears in the reference list in any case, but to confirm exactly which I need, I sometimes specify this alongside it, using one of the following abbreviations:  [м.] = masculine, [ж.] = feminine, [ср.] = neuter;  [ед.] = singular and [множ.] = plural.

   Adjectives are a particular problem in Slavic, because some languages (Bulgarian and Macedonian) only have short form adjectives, while others (Ukrainian) have only adjectives derived from the Proto‑Slavic long forms.  In most cases I will only be comparing phonetically the roots of these adjectives, but for now please record them in the same morphological form as in the reference list.

   In some cases the reference list contains two words separated by a comma [,], e.g. брат, братья.  In these cases I need both forms, so please try to record both pronunciations.

   In some cases I need a word root which in some dialects may not really be used on its own, but only be common as part of a compound.  I write such roots with a hyphen [], e.g. благо‑ or добро‑.  To help your informant you can mention some compounds that include this root, such as благодать or благороден, but for the actual recording please try to get the informant to pronounce only this root благо‑, not the rest of the compound word.

   All words need to be pronounced in isolation, without any preceding or following words.  If speakers do add some other word or suffix, please ask them to repeat the word in the reference list, on its own.  (It is especially common for informants to add the definite articles in languages that have one;  for Russian this at least will not be a problem!)

   If your informant’s normal form of a word adds a diminutive suffix to the root, please try to record him or her also pronouncing the bare root without a diminutive, if that is at least possible in his or her dialect;  if it really isn’t, please note this alongside the word in the reference list.

   To help in eliciting the correct word, before and/or after the word I need recorded I put some other words to give the context — in small letters and only in grey ink.  This is just to help the informant recall the exact form of that word in a real sentence context.  I do not want to have those ‘context’ words recorded, however.  Please ask your informant only to pronounce the main reference word, not the full context. 

   Many informants have a habit of saying the word ‘and’ just before the last word on the page!  In the sound file this can be difficult to separate from the pronunciation of the following word, so if your informant does this, please ask him or her to repeat the word without ‘and’ in front.

   Finally, please make sure your informant doesn’t read the list too fast!  There has to be at least a short pause and silence in the recording between each of the words when pronounced.


*     *     *     *     *


If you have any doubts, you can contact me by email [paulheggarty AT], or on Skype [Paul Heggarty], or I can call you on a normal telephone if you give me a number. 

Please let me know too if you find that I have made any errors in my Russian in the reference word list.

Thanks once again for your patience in reading these instructions, and for your willingness to help.  All the best for your fieldwork trip!