IT feels like the Gaelic census results have been as long coming as this General Election.

Both have finally been announced, and the language headlines are that 12,000 more people say they speak Gaelic – taking us up to 70,000 – and that up to 130,000 have “Gaelic skills” (read, write or understand) which is a large increase.

On the face of it, these numbers should be a relief to anyone who is working in Gaelic development, or has done so over the past 20 years. Many, many people have put blood, sweat and tears into Gaelic revitalisation – often with little thanks – and their efforts should not go unrecognised.

The next logical step is to drill into the data and work out how many fluent speakers, intermediate or early-stage learners we have in each area. We can work out a plan from there to meet community needs and grow language use.

Except we can’t. Because the question was a binary “yes or no” to the four skills – speak, read, write and understand. We have literally no idea whether our new speakers are able to converse at a high level, or whether they have recently downloaded Duolingo. Not a clue.

So, if we don’t know what level speakers or skill holders are at, what do we know? Well, we know where speakers are based. We have age and location relative to the answers. Broadly, numbers are rising in the cities and dropping in the Gaelic heartlands. Which makes the figures feel less positive the longer you look at them.

READ MORE: What does the census tell us about Scottish and British identity?

There are always exceptions, and Sleat in Skye is one bucking the trend. It is the location of Sabhal Mòr Ostaig, so there would be serious questions to be asked if it wasn’t having a positive impact on the community after 50 years. Amazing what targeted and consistent funding (poor though it may be) can do, isn’t it?

At the other end of the spectrum, an early slicing of the numbers shows that for the first time, Gaelic speakers are in the minority in the Western Isles. If ever there was a canary in the mine, that’s it. And it’s coughing.

We could have got the answer to the skill level conundrum – even in a crude form. The question about English offered options relating to ability. The next question – relating to Gaelic and Scots – did not. It means that for our two threatened indigenous Scottish languages, it didn’t occur to anyone that slightly more granular information would be helpful. Or did it?

Census questions don’t appear out of thin air. Someone makes a choice about what is asked.

The prognosis for Gaelic as a community language is increasingly poor, but that doesn’t look good politically. And as recent funding cuts have shown, there is no great appetite to fund Gaelic development appropriately at a government level. If the data gathered is loose enough to be interpreted as being “better than expected” and “encouraging”, then the pressure comes off, and the additional funding requests become easier to refuse.

The cynical among us might suggest that was the line being walked by the Gaelic-speaking Deputy First Minister Kate Forbes this week as she told the Education Committee that she preferred the word “urgency” to “crisis”. We all prefer the word urgency, but I’m not sure what a collapse in daily language use in our vernacular communities constitutes if not a crisis. To frame it any other way is a cop-out.

READ MORE: New book proposes 'alternative path' for Scottish independence

Duolingo is mentioned a lot. It has had a huge impact on the awareness of Gaelic in Scotland and has given people an easily accessible way to learn some. There is nothing wrong with Duolingo and I’m glad the Gaelic version exists. At the same time, the percentage of people who will become functionally fluent off the back of it is surely vanishingly small.

Duolingo aims to offer courses up to B2 – the level at which you could get a job in that language. It doesn’t say how many hours you would need to put in, but I reckon that the average bear would need a streak of a good few years and some supporting resources.

My cynicism aside, the fact that Duolingo has a reference point and assesses its courses based on that reference point puts it streets ahead of our census results. It is also at pains to point out that where you need to get to with a language depends entirely on what you want to do with it.

What people want to do with the language leads us to the elephant in the room – what constitutes a Gaelic speaker? That is what the questions being asked around the census results boil down to. And the topic is so fraught that few are willing to ask it out loud, fearing a pile on suggesting we are somehow elitist or racist, which is about as helpful as, well, this census question.

Instead, we dance around the issue, hoping that it will somehow answer itself. The answer from the census results is “anyone with some form of Gaelic”. That also feels like a cop-out.

Is there a better answer? Is a Gaelic speaker someone who grew up in a Gaelic-speaking community with Gaelic in the home? Is it someone like me who grew up in Edinburgh in Gaelic medium education and achieved fluency through using Gaelic at work? Is it someone who is doing the Cùrsa Inntrigidh at Sabhal Mòr? Someone who did six weeks of Ulpan in 2006 in Cumbernauld? Is it someone with a 3000-day streak on Duolingo? Is it someone who went through Gaelic medium primary but hasn’t spoken since? Someone who did Gaelic all the way through uni?

READ MORE: Gerry Hassan: Continuity politics will not be enough, whether it's Labour or SNP

Is it all of them? I’m not sure I care. The critical thing is that they open their mouths and give the “speaker” part some credence.

The only way to save Gaelic as a community language in its heartland areas is to have people who can speak it as a daily language. People who speak out loud in public. It’s irrelevant whether they are native speakers or fluent learners.

Without audible speech, the house of cards falls. Teaching it in school becomes little more than an academic exercise with some history thrown in. Learners have no-one to converse with save each other. The dialects diminish, the traditions fall away and all you have left are the words. Even then, the wealth of vocabulary disappears (more so than it has already), and with it the richness. And, some might argue, much of the point.

I’d go further than saying that the census question wasn’t nuanced enough. I don’t think it’s the right question full stop. Rather than focusing on ability, the question should be: “How often do you speak Gaelic, and to how many people?”

I don’t speak Gaelic everyday, and when I do it’s normally to the same three or four people. Moving out of that comfort zone is incredibly hard when it’s become the norm.

I think the answer to that question – how often and how many – would give us a much clearer picture of where we are at. It’s a question we don’t want to ask – each other or ourselves – because we know the answer.

But to make any progress increasing daily use in places like Tiree, we must ask ourselves that question. Our results, 230 speakers, are being reported as better than expected, suggesting we’ve only seen a net loss of 10 speakers since the last census. If they could all stand up and make themselves known, that would be great. If they would open their mouths and speak, it would be even better.

The painful truth is that gaining 20 speakers who only talk to an owl, while we stand silently at the graves of our tradition bearers, isn’t a good news story. We’re losing far more than we’re gaining.