I get requests for help on occasion from people who want me to critique their drawings over the Internet. As I've spent a number of years refining my ability to see and draw, and have also spent years teaching, it's very difficult for me to impart principles of drawing that require long study to acquire an understanding of, and even longer to implement in an effective way. It has been my personal experience that a mental knowledge of certain principles is usually moving quicker than the ability to execute them. So realistically, the best advice I can give is to seek a proper education and spend the time refining your craft. However, I understand that there are people in certain situations where such a commitment in time, finances, and other difficulties aren't able to pursue the education they need, and so look for other alternatives.
So I thought it would be helpful for those that have contacted me and possibly for others, if I spent a little time outlining how to start a drawing, because obviously a solid start is a foundation for a solid finish. People seem to naturally love detail and finish, but without an appropriate underpinning and accurate relationship of those details to "the whole" the drawing soon falls apart. Those familiar with Sargent's working methods as they have been written about detail his discipline in returning to correct a poor foundation in his work, as each large area should be accurately placed to receive subsequent smaller details. Which brings me to the first working principal, and that is working from general to specific. So then, how to start? Start by dictating the longest measurement of your subject, and be conscious of what you're choosing to include on the page. Nothing marks an amateur draftsman like an inability to capture feet and other extremities without running off the page. Having this measurement established will force you to make all other decisions relative to it from the start, including whether the longest opposing measurement will also fit on the page. Now, procedurally the most common mistake people make in beginning a drawing is that they start drawing. All those gorgeous details. A drawing in it's early stages should be no more than a series of indicating lines. At its most basic, those indicating lines are broken off of an all encompassing "envelope". This envelope establishes our extreme limits of proportion. The main goal in the beginning is always to capture the most information with the least amount of effort, or to put it simply, line economy. By adhering to this, we eliminate the need to move multitudes of information as we encounter errors, and refine our ability to see more final accuracy in the early stages of the drawing.
Working large to small is something that many people are probably already familiar with, but allow me to clarify it a step further. There is a large difference in stating something in a general way, as opposed to a generic way. I think the example labeled “generic" will suffice to explain this. In the general version or the enveloped version, you will see that even though the information stated is stated broadly, that information is specifically related, and so sets the stage for accurate further refinement. And the generic version, it is just that, generic. There's a semblance of the original subject but we lack a sure foundation of accuracy. Now comes the question of further refinement. What next? If my watchwords are general to specific or large to small, it doesn't make much sense for me to start drawing noses and toes after I've established my envelope. So consider it like a block of marble. What are the next cuts in this block that I can make to start to view the more true nature of my subject as soon as possible? Being a figure the dominant rhythm of the pose is the first thing to consider. What are the large opposing thrusts? How is the figure balanced in relationship to those thrusts? Are the angles of those thrusts reflected in other areas? If I carry the lines of those angles through the figure what would they relate to?
As you proceed through the drawing, there is a tendency for the mind to slide along a scale. On one end of the scale is knowledge, for instance anatomy, on the other side is naïveté. Neither extreme is particularly bad but it's better to be somewhere in the middle. Solely focusing on the abstract shapes that make up the figure in a naïve way might lead to some deficiencies in the long run understanding of how a figure is constructed, whereas infusing extreme knowledge of anatomy into the figure might end up giving an unintended result as well, possibly overpowering the unified grace of the pose.
Ultimately the greatest importance in the beginning of the drawing is always large relationships, and the relationships of all the parts to the whole subject or "the whole”. Reduce round areas into series of straight lines. If it requires multiple lines at multiple angles as you search for the drawing, immediately clean up the lines and retain the one that is most correct. Keep your visual space clean. In the example images overlaid on a Charles Bargue Figure Drawing, you will see the angles of relation that you could see, and a cleaned up version which might look like your drawing after 40 minutes. You don’t necessarily have to physically draw through the figure all the relational lines but seeing them is what is important.
At a certain point in your development you will be able to skip steps, infuse variety and idealism, but as Harold Speed has said in his excellent book on drawing “The Practice and Science of Drawing”
"…In connection with academic drawing, it is eminently necessary for the student to train his eye accurately to observe the forms of things by the most painstaking of drawings. In these school drawings, feeling need not be considered, but only cold accuracy. In the same way a singer trains himself to sing scales, giving every note exactly the same weight, and preserving a mechanical time throughout, so that every note of his voice may be accurately under his control and be equal to the subtlest variations he may afterwards want to infuse into it at the dictates of feeling. For how can the draughtsman who does not know how to draw accurately the cold commonplace view of an object hope to give expression to the subtle differences presented by the same thing seen under the excitement of strong feeling?"