Inkwell: IELTS Academic Reading practice test

There is no lead in a pencil and there never has been. The writing material is graphite, a form of carbon whose atoms lie in sheets that slide over one another readily, which is why it leaves a mark on paper and why it feels slippery between the fingers. The confusion is old and comes from a mistake of identification: when an extraordinary deposit was found in Cumberland in the 1560s, the black substance was taken for a variety of lead ore and named accordingly, and the name stayed long after the chemistry was understood.

That deposit, at Borrowdale, is the reason the pencil exists in the form it does. It was the only large body of solid graphite pure enough to be sawn into sticks and used directly ever found anywhere; every other source in the world required the material to be ground and reconstituted. The sticks marked paper beautifully and marked the hand equally well, so they were first wrapped in string that could be unwound as the point wore down, and later enclosed in a wooden case. The English had a monopoly on the material for two centuries, and the mine was worked only a few weeks a year, guarded by armed men and deliberately flooded between workings to prevent theft.

The monopoly ended not with a discovery but with a war. Cut off from English graphite during the conflicts of the 1790s, the French government asked an officer named Nicolas-Jacques Conté to find a substitute, and within a few days he had one. Powdered graphite of ordinary quality, mixed with clay, moulded into rods and fired in a kiln, produced a usable core from material that had been useless before. It also produced something the Borrowdale sticks never could: a range. The more clay in the mixture, the harder the core and the paler the line it leaves; the more graphite, the softer and the blacker. Every grading system in use since, with its H for hard and B for black, descends directly from a wartime expedient.

Modern manufacture would be recognisable to a worker of the 1860s. A slat of wood about the length of a pencil and the width of several is machined with parallel grooves; the cores are laid into the grooves; a second grooved slat is glued on top, making a sandwich; and when the glue has set the sandwich is cut lengthwise and shaped, so that a single pair of slats yields eight or nine finished pencils at once. The painting, the ferrule and the eraser are afterthoughts added in the nineteenth century, and the hexagonal cross section exists because a round pencil rolls off a sloping desk.

The wood is a more demanding requirement than it appears. It has to be soft enough to be cut by a small blade held in the hand, without splintering, and it must cut as cleanly across the grain as along it, since a sharpener does both. Very few timbers manage this. The industry used red cedar from the eastern United States until the supply was exhausted, at which point it moved to incense cedar from California, which remains the standard; substitutes have been tried repeatedly and most of them tear.

Given how much has been invented since, it is worth asking why the pencil is still made in billions. It writes upside down, at any angle, in extreme cold, on damp paper and in a vacuum, none of which a ballpoint pen reliably manages. It requires no power, no cartridge and no maintenance beyond a blade. Its mark can be removed and, if it is not removed, will still be legible in three hundred years, because graphite is chemically inert in a way that no ink is. For a certain range of tasks nothing has ever improved on it, and the range includes several that matter.

There is a well-known essay which observes that no single person on earth knows how to make one. The claim sounds like a rhetorical flourish and survives inspection. Making a pencil requires cedar, which requires forestry, saws and the steel to make saws; graphite from one continent and clay from another; a kiln, and the fuel and refractory materials a kiln needs; glue, lacquer, brass for the ferrule and rubber for the eraser, each with its own supply chain reaching back through mining, chemistry and transport. The person operating the machine that cuts the slats knows nothing about vulcanising rubber, and does not need to.

That is the pencil's real interest. It is among the cheapest manufactured objects in existence, costing less than a few minutes of anybody's labour, and it is the product of a coordination that nobody designed and nobody oversees. The person who wants one does not have to know any of this, and would not be able to obtain one if they did, because knowing how a pencil is made is not the same as being able to make one, and no amount of study closes the gap.

There are somewhere around seven thousand languages in use, and the distribution of speakers among them is extraordinarily unequal: a few dozen account for the great majority of humanity, while roughly half have fewer than ten thousand speakers each and several hundred have fewer than a hundred. A projection that circulates widely holds that half of the total will have no remaining speakers by the end of this century. The figure rests on assumptions that can be argued with, but the direction is not in doubt, and the rate at which small languages are ceasing to be transmitted is without precedent in the historical record.

It is worth being precise about how a language ends, because the popular image is wrong. It rarely happens through the death of its speakers. It happens when a generation of parents who speak it fluently decide, consciously or otherwise, not to use it with their children, and they decide that because they can see clearly that their children's schooling, employment and safety will depend on a different language. Within two generations the transmission has stopped; the last fluent speaker dies perhaps sixty years later, and it is that death that gets reported. By then the decisive events are two generations in the past.

What exactly is lost is a question the field has not always answered honestly. The strongest version of the claim, that each language contains a distinct way of perceiving the world which disappears with it, is not supported by the evidence; the experimental work on linguistic relativity finds effects that are real, consistent and small, concerning things like the speed of colour discrimination or the ease of certain spatial tasks. Two weaker claims are much better founded. The first is that every language is a data point about what human grammar can do, and since the shape of linguistic theory depends on the range of the sample, losing the unusual cases distorts what we believe to be possible. The second is that a great deal of specific knowledge, about plants, animals, weather and navigation, exists only in the vocabulary of small languages and in nobody's written records.

Ingrid Solberg, who has worked on language documentation for thirty years, raises an objection from inside the field that deserves a hearing. Almost all of the advocacy, she notes, comes from people who will not bear any of the cost. A linguist urging a community to maintain its language is asking parents to accept, on their children's behalf, a measurable disadvantage in schooling and employment, in exchange for a benefit that accrues largely to scholarship and to an abstraction called human heritage. Speakers who shift to a dominant language are not, in general, confused about their own interests. They have more information about their situation than any visiting researcher, and they are frequently making the choice that a reasonable person in their position would make.

This seems to me both correct and incomplete. The choice is real, and treating speakers as passive victims of historical forces is both false and patronising. But a choice made under conditions somebody else arranged is not quite the free choice the argument requires. Where education is available in one language only, where the state conducts its business in one language only, and where speaking the other has been punished within living memory, the decision not to transmit is a response to a set of costs that were imposed rather than discovered. Removing those costs is not the same as urging people to bear them, and it is what most of the successful cases have actually involved.

Those cases are worth looking at, because they are unusually consistent about what works. The most effective single intervention has been the immersion preschool, in which very young children spend their days with fluent elders and no other language is used; the model developed in New Zealand in the 1980s and was adopted in Hawaii, and in both places it produced a generation of speakers where transmission had effectively stopped. Where numbers have risen for a language with more speakers, as in Wales, the change followed a shift in education policy rather than a campaign of persuasion. The common element is not enthusiasm. It is the creation of situations in which the language is necessary rather than merely permitted, because a language that is only ever optional becomes the one you use with your grandmother.

Documentation is a different activity from revitalisation and the two are often confused. A recorded corpus with a grammar and a dictionary is not a living language and cannot be turned into one by scholarship; what it preserves is a record. That said, the record is not nothing. At least one community in North America has used nineteenth-century archival material, including documents written by speakers themselves, as the basis for reconstructing a language that had had no speakers for a century, and there are now children being raised in it. That outcome depended on an archive, on a linguist from the community, and on a decision taken by the community rather than for it.

Which is where the argument ought to end. The question is not whether the world's languages should be preserved, which is a question nobody outside a community is entitled to answer on its behalf. It is who decides, and under what conditions the deciding happens. A community with schools, employment, official recognition and a functioning archive that chooses to shift has made a choice. A community without any of those has been presented with a situation. The useful work for everybody else lies in narrowing the difference between the two, and then accepting whatever is decided.