On the Supposed Universality of Draconic Speech

The assertion that every Dragon speaks every language has enjoyed a success quite independent of the number of Dragons anyone has heard speaking. Its advocates commonly retreat, when pressed, to the suggestion that Dragons conceal their accomplishments out of contempt for us. This explanation confers upon the silent Dragon both the faculty under dispute and a motive for withholding evidence of it; I congratulate its inventors upon having made themselves so completely unnecessary to an investigation. Most Dragons never produce sapient speech. A very few do, and their facility in acquiring additional languages is sufficiently remarkable to deserve an account less flattering to our appetite for marvels.

I must first distinguish speech from several performances which visitors to Merewick habitually mistake for it. A Dragon may reproduce a sentence with excellent fidelity, including the cough of the person from whom it learned it, without employing any part of that sentence independently. One of the Dragons allied with the town announces that the west gate is unlocked whenever the carter’s wagon comes up the hill. The gate may be locked, open, or undergoing repair; the announcement remains equally confident. Senn, the town’s go-between with its Dragons, demonstrated this to a visiting philologist, who had already described the Dragon as a conversationalist. I asked him whether he proposed to extend the same courtesy to a defective station clock. He subsequently amended his description to “a conversationalist of uncertain veracity”, which preserved his opinion at the expense of his subject’s character.

Understanding requires separate examination. Many Dragons respond appropriately to spoken requests, and some distinguish alterations which cannot be explained by familiarity with a single sound. If asked to bring the smaller of two unfamiliar objects, such a Dragon may choose correctly even when their positions are exchanged and the speaker stands concealed. These results establish something worth knowing, though nothing so large as that the Dragon could describe the choice, explain the relation of size, or understand a similar comparison concerning debts. I have repeatedly found comprehension exceeding expression, sometimes by an extraordinary margin. The townswoman who says that her Dragon neighbour understands more than he can tell her may be reporting a sound observation, although I should still like to see how she obtained it.

Among the few Dragons whose speech I have examined adequately, a curious regularity appears. Once an individual can converse in one language, additional languages seem to offer comparatively little resistance. Acquisition still requires exposure: a Dragon does not understand the speech of a country merely because a geographer has coloured that country upon a map. Yet after sufficient acquaintance with its speakers, the Dragon commonly attains much the same degree of competence that it possesses in its first language. The same powers of explanation appear, together with the same peculiar hesitations. An idiom may be missing; the underlying difficulty of speaking scarcely changes. Our evidence remains small, and encounters with genuinely unfamiliar languages are difficult to arrange without introducing an interpreter who contaminates the experiment by being helpful.

The popular account has therefore enlarged an uncommon faculty into a universal possession, then abolished the trouble of learning it. I suspect travellers have assisted this process quite innocently. A Dragon encountered beside an old trade road may have heard several languages for longer than the traveller’s nation has possessed its present name. Its answer in a foreign tongue appears miraculous chiefly because nobody has enquired into its previous company. Nevertheless, accumulated exposure cannot explain the whole matter: records kept at Merewick show speaking Dragons acquiring newly introduced tongues far more readily than the townspeople who taught them, despite receiving fewer hours of direct instruction. These records are imperfect, but they contain enough repeated observation to make dismissal as foolish as credulity.

Neither general intelligence nor the construction of the vocal organs satisfactorily explains why speaking individuals are so rare. I have watched a Dragon solve a complicated fastening which defeated a competent machinist, then fail through weeks of patient instruction to use two familiar words in a fresh combination. Conversely, one of the less resourceful Dragons allied with the settlement at Harrowglass could give a tolerably clear account of an absent neighbour’s intentions. Silent Dragons may reproduce the sounds required for speech; articulate Dragons possess no corresponding uniformity of tongue or palate that our examinations have detected. There may be relevant differences beyond our present means of inspection, and I will not convert the limitations of an examination into proof that nothing remains to be found.

The difficulty appears to concern the preparation of experience for expression. Our own speech depends upon extracting something from the circumstances in which we encounter it and making that portion available for other uses. We can discuss a promise without naming its maker, or a doorway without deciding what approaches through it. Draconic attention seems to preserve connections which we readily set aside. An object’s approach, its former uses and the expectation it arouses may remain involved in what the Dragon apprehends as that object. I offer this as an inference from behaviour and testimony. We cannot enter a Dragon’s perception, and a fluent individual describing it has already performed the very operation we are trying to understand.

My most instructive conversation upon the subject occurred at Merewick with a Dragon called Ash-of-the-Weir, whose command of Caldric I had previously tested and whose reputation for disagreeableness Senn considered deserved. Ash was resting against the sun-warmed wall of the tithe barn when I asked whether she had expected Senn’s predecessor as go-between to return after his retirement. She answered, after a considerable interval, “His coming remained.” I requested a less figurative reply. “You ask,” she said, and stopped long enough for me to repeat the question. She struck the paving with one claw. “I am answering. Your second asking is now also here.” Senn advised me to wait. I replied that I was perfectly capable of waiting, an assertion whose experimental refutation had just occupied less than a minute.

Ash eventually explained that the old go-between’s approach had included a dragging boot, the smell of the leather dressing he used, and a particular vibration transmitted through the timber bridge he crossed to reach her. After his departure, portions of this familiar occurrence continued. Senn’s apprentice wore his old coat; a loose plank on the bridge supplied something resembling his uneven step. To my question whether she had therefore mistaken the apprentice for him, she replied, “No. That one was himself. The coming was unfinished.” Her emphasis fell heavily upon “was”, while “himself” followed in a quick, almost incidental murmur. I initially took this for a distinction between identity and expectation, but she rejected my formulation twice. I record it here because the rejected explanation is tempting, and because I spent an embarrassing quarter of an hour attempting to obtain her consent to it.

I then asked her to suppose that the old go-between had returned by another road, wearing different clothes. She understood the instruction readily enough to complain of it. “You remove everything by which the question reaches me. Then you keep him.” I insisted that the man could survive a change of coat. “Of course,” she answered promptly, and for once with perfectly ordinary emphasis, as though the point hardly needed making. Senn looked away. The obstacle concerned my demand that she form and communicate an account of his return while excluding the circumstances through which that return acquired its particular significance for her. She could accomplish this, but required time. Her eventual answer was simply that she would recognise him and would be surprised by his approach.

At this point I made the error of asking Senn whether Ash always found hypothetical questions difficult, as though our interlocutor had become an exhibit upon the conclusion of her last sentence. Ash interrupted in Dwarven, which Senn speaks with greater ease than Caldric: “He is making me smaller so that his question fits.” I disputed the accusation. Simplification, I explained, was necessary to determine which difficulty we were discussing. She answered, still in Dwarven, that I had simplified her answer before allowing her to finish it. Upon consulting my notes, I found that she was correct. I apologised and invited her to specify what I had omitted. This produced the most useful part of the interview, though it also prolonged it sufficiently to cost me my train.

Her account suggested that listening allowed her to follow distinctions supplied by another mind, whereas speaking obliged her to establish those distinctions herself. “When you speak, I can go where you have divided,” she said. Then, after a pause: “For speaking, I must find where to begin the dividing.” She disowned “divided” as an adequate word, but accepted it provisionally after rejecting several alternatives of mine. This accords with the wider observation that Dragons often understand statements which they cannot independently formulate. It also offers a possible explanation for the multilingual speaker: once the difficult conceptual preparation has been achieved, attaching the resulting distinctions to another vocabulary and grammar may demand relatively little additional effort. I consider this explanation promising, though I do not know whether it is correct.

Ash’s passage into Dwarven did not accelerate her speech: in both languages she paused within clauses at places where a Human listener expects continuity, and hurried across boundaries where we ordinarily allow a breath. Her pronunciation was competent; her emphasis could make a subordinate word sound like the purpose of an entire discourse. These features persist even in Dragons with generations of conversational practice. Some pauses plainly concern the effort she described, though others may reflect habits of attention we have yet to identify. An investigator who fills every silence ends by recording chiefly his own speech, so I now ask Merewick’s interpreters to allow a Dragon to signal completion, and have adopted the same precaution myself, with results more satisfactory than my early confidence warranted.

The rare speaking Dragon consequently gives us access to a difficulty which the silent majority cannot explain. We should resist making that witness representative in every respect: the capacity to tell us about Draconic thought may itself accompany an unusual manner of thinking. Even so, the evidence supports a limited conclusion. Most Dragons cannot converse in sapient language; many understand appreciably more than they can express; the few fluent speakers appear able to acquire further languages with an ease deserving systematic study. Ash has agreed to further interviews, provided Senn attends and I bring fewer questions. I argued that reducing their number might leave important matters unexplored. She replied that I had yet to discover how many questions there were in the first one, and I have accordingly revised the programme.