The incompetence of most job talks seems astonishing and depressing until one remembers just how badly conceived and delivered conference talks tend to be. If oratorical savvy is in such short supply, then the inventive job hunter has a shining opportunity to better his or her chances. To give an impressive talk, you need to understand 1) the basics of effective oral presentation; 2) the nature of talking to a whole department, not just to fellow specialists; and 3) tactics specific to the type of institution. Do these three things well and you will tilt the odds heavily in your favor.
An article is not a talk. A thesis chapter (even shortened) is not a talk. The kind of language that projects your masterful arguments in written prose is indigestible sludge when spoken. Learning through your ears and learning through your eyes are different, so you have to adapt your material to different mental processes. For the ears, you want an exaggeratedly clear roadmap and short sentences. Close to the beginning, you want to say "I shall address three problems: a, b, and c." You make sure that you mention a and b and c at the beginnings of their respective sections, and you remind us at the end that you have solved these three problems. Whether you solve problems or answer questions or trace an unusual development, get it up there in plain English, and remind us from time to time where you are in the trajectory of your argument.
The average length of your spoken sentences should be shorter than that of your written prose. Sure, a few may be longish, but anything complicated should be broken into two or more short sentences. The John Gardner rule will help you here. Imagine that you are diagramming a sentence. It has three basic parts: Subject, Verb, and Object. All the rest, clauses and phrases, are variously attached to one of those three. His rule: never complicate more than one of them. If you have a complicated subject with lead-in phrases, modifying phrases, and maybe a list or two, then be sure your verb and object are simple. Use that rule to help you see what needs to be simplified. When I teach Ph.D.-level technical writing, that rule works wonders. Incomprehensible subjects such as vapor-deposition chemistry and opaque risk analysis start to be understood by the class after a John Gardner-style revision.
Cut the jargon. By jargon, I mean the buzzwords that you do not use in everyday speech: imbricate, reinscribe, transhistoric valences, aesthetic reciprocity, perform (when you are not talking about acting on stage). Very occasionally, you will need a technical term because we have no ordinary equivalent, but that is rare. Imagine describing your subject to your mother and avoid words she would not understand. You might set yourself a limit: no more than one or at most two buzzwords per page. That way you can use a few to signal that you talk the talk. Remember, though, that not all members of a department belong to your same knowledge community and rely on your favorite theorists.
Print your talk in 16-point font so it can lie flat on the podium but you can still read it easily even if there is no lectern. Mark the most important word or phrase in each sentence in boldface caps. Read the talk aloud (not silently) 8 or 10 times, so that you have it three-quarters memorized. That way you can deliver it while looking at your audience and making eye contact. You need only look down occasionally to remind yourself of what comes next. Those boldfaced words will lead your eye right to where you need to be. People who mumble downward at their text make a bad impression. If you are not one of the gifted few who can speak extempore, then at least memorize and practice so that you interact throughout with your listeners.
You need to deliver your ideas with expression and emphasis. Wave your arms a bit. Don’t look like a frozen-faced robot. Smile. Be exaggeratedly ironic. Show some emotion. You have to project real excitement over your insight or discovery or no one else will feel any. Once, but only once, in the talk, say “What I really love about x is….” People want to know that you love your work. Exhibit some bounce. Dull, mumbling, inaudible teachers do not help a department build enrollments.
All that I have said so far applies to conference talks. When you are addressing a department, you need to meet some other needs. You are trying to show them that you will be dynamite in the classroom and you must appeal beyond your subspecialty. Start with a general point that would be of interest to a broad audience, not a narrow question that matters only to fellow specialists. Even more important, you want something flashy and exciting, something that will make your listeners sit up and feel glad that that they bothered to attend. If you can come up with some relevant anecdote or fact that is funny or grotesque or puzzling or even disgusting, use it to get things off to a lively start. Deliver this with relish and then give your road map and make your serious argument.
Similarly, when ending, try to open up to broader implications or applications. Your main argument may be very field-specific, but you want to end on a note that suggests to people in other departmental fields that what you have found, or your methodology, or your new slant, might be useful to some aspect of their work too. That makes you look like a good investment as a new colleague.
How you pitch your argument will depend on the institution interviewing you. A few very elite universities believe in intellectual hazing, and will think you a wimp if you don’t try to cow them with a brilliant, jargon-laden analysis. If you come from such a department, you presumably know how this game is played and can prepare for it, and only people from that kind of university are likely to be invited for a campus visit. Most departments are more welcoming and friendly; they want a good colleague with whom exchanging ideas will be fun, not someone determined to bully them intellectually in every exchange. You may meet someone who emerged from that kind of training in the audience, and that person may try to give you a hard time in the Q and A, but most of your potential colleagues don’t feel the need to prove themselves that way.
Hence, for most job talks, I recommend using as little jargon and article-style argument as possible. You show your intellectual substance in the Q and A. Listen carefully to each question, and try to answer crisply and directly. Give a mock job talk to your fellow job hunters and faculty friends and make notes on the Q and A, so you know what sort of questions you are likely to draw. If you prepare in this fashion, you will do yourself credit and come out way ahead of most of your rivals.
Kathryn Hume is the Edwin Erle Sparks Professor of English at Pennsylvania State University. She is the author of Surviving Your Academic Job Hunt: Advice for Humanities Ph.D.s  (Palgrave Macmillan).