It can inspire contempt as mere "slang" or awe as the language of Robert Burns and Allan Ramsay.

Scots has long provoked very different reactions - not least politically. Some diehard unionists detect a nationalist agenda in new efforts to support the language.

But does speaking Scots really make you more likely to support independence?

Voting patterns in areas where Scots is most spoken actually suggest the reverse.

Scots is most widely spoken in the northern isles of Orkney and Shetland and in the north-east council areas of Aberdeenshire and Moray.

Fully 49 per cent of Shetlanders and 48 per cent of Aberdeenshire residents told the 2011 census they knew Scots, compared with a national average of 30 per cent.

But come 2014 and support for independence in these areas proved to be far lower than knowledge of the language.

Only 36 per cent of Shetlanders voted Yes, well below the national average of 45 per cent.

In Orkney, according to the census, 40 per cent of residents say they speak Scots. Only 32 per cent of voters backed independence, the lowest figure in the country.

In fact, all four areas with the highest levels of Scots-speaking residents had below average support for independence.

So what about the four areas of Scotland which voted Yes? Where did they rank on prevalence of the Scots language?

Dundee was the most pro-independence place in the country last year, with a Yes vote of 57 per cent. However, only 31 per cent of Dundonians told the census they speak Scots.

That is just below the national average of 30 per cent.

But Dundee, in fact, ranks 17th out of the 32 councils on this measure, just below average.

The other three council areas that backed independence were Glasgow, North Lanarkshire and West Dunbartonshire. All three are firmly below average on Scots speakers. In Glasgow, the so-called Yes City, only 25 per cent of people said they had the language.

Several areas with very low level of Scots - such as East Renfrewshire, Edinburgh, Dumfries and Galloway and Scottish Borders - voted firmly No.

Meanwhile, the Western Isles - with the highest prevalence of Gaelic - has a very low figure for knowledge of Scots. Its support for independence was slightly above average at 47 per cent.

The Scots language - perhaps with its strong regional variations often underlining local rather than national identity - doesn't seem to correlate with support for independence.

But what about areas with high levels of Scottish identity reported in the census? Are they more likely to vote Yes?

Here too the evidence of a strong determining link is hard to prove.

In the 2011 census the two places that recorded the highest levels of Scottish-only national identity in the 2011 census were West Dunbartonshire and North Lanarkshire.

They both scored at 72 per cent compared with a national average of 62 per cent. Inverclyde, which voted No by a whisker, ranked fourth in a league table of Scottish only national identity.

However, the Yes cities of Glasgow and Dundee did not record particularly high levels of Scottish-only identity in 2011 census.

Dundee was above average at 65 per cent.

Glasgow was at the national average of 62 per cent. Yet so too was Orkney, the place where support for independence was lowest of all.

Only in Edinburgh does "Scottish-only" identity fall below half of the population - at 49 per cent. Support for independence in the capital was the seventh lowest in the country, at just 39 per cent.