A tradition captures the imagination of contemporary makers

Ultimate Small Shop

Start a Woodworking Business From Home

Get Instant Access

by Jeremy Singley

Contemporary Southern Chairmakers


Back bow



Optional pegs help resist shock.

Tapers on legs and posts ensure gap-free fits even if sizes are slightly off. Tapered foot design allows length adjustments after assembly, unlike ball-foot and other designs. Similarly, tapered back spindles allow trimming after assembly, allowing for minor variations in back curves and spindle lengths.

Fig. 1: Windsor engineering

Optional bracing spindles lengthen triangle's base.

Forward position of back bow in seat creates a triangle with the center spindles, swinging the crest rail's load downward to convert backward shear force on center spindles into the compressive force wood sustains best. Wedged spindles are a weakness because load tends to force spindles through bow.


Back bow

^ Screw with plugged head


- Arm post


Back bow

Splayed legs triangulate with rungs and seat, but rungs are still the design's weakest point. Weight of sitter tends to spread legs. Windsor-makers overcame this drawback, with only partial success, by shrinking undercut holes in green legs onto dry ball-end rungs. Rungs were often made slightly over-long so the undercarriage was 'pre-stressed.'

Forward lean of arm post triangulates with backward lean of back bow. This arm-post joint is tapered like leg joints, but sitters too often pull up on arm, and the backward swing of the back bow spindles under load adds to this upward force, making the tapers a liability.

Tapered socket joints of legs tighten under sitter's weight.

Undercut socket made with brace and spoon bit.

Chair has more parts than a standard mortise-and-tenon chair, but parts are cheaper to make: Legs, spindles and back bow are rived, or split, from green wood, which is faster than sawing and ensures continuous grain lines for maximum strength. Curved parts are generally bent, saving material; legs are lathe-turned green, allowing heavy cuts. Seat is sawn from solid plank, saving glue-up time.

110 The Best of Fine Woodworking

From l-'ine Woodworking magazine (July 1988) 71:32-38

The work of Michael Dunbar, one of the leading makers in the rebirth of the handmade Windsor, demonstrates some of the adaptability of the concept: a shaped plank seat bored with sockets to accept an undercarriage and superstructure

Had George Nakashima taken his work less personally, my chairmaking career might never have begun. A woman who bought two of his very first captain's chairs, years before, feared to return them to their source tor repairs because she knew the famed contemporary-style woodworker was known sometimes to revoke ownership of work he thought the owner had abused. Instead, she dropped them off, in pieces, at my daddy's shop one serendipitous day some 17 years ago. I was fascinated. The chairs were all of oiled walnut and showed signs of being handmade. The spindles bore heavy spokeshave marks like facets on a cut diamond, and the joints were precisely turned and pegged in place. But when the chairs were restored, I found them uncomfortable to sit in. My attempts to better them ensued, and I began developing my own versions of sleek all-hardwood spindle-back chairs.

It was years before I began to realize I was developing variations of Windsor chairs, a form that certainly didn't start in my lifetime or Nakashima's. My ignorance of the style's pedigree could only be a liability to my work, perhaps bordering on arrogance. I hit die books and talked to people who knew, and I was happily humbled.

The Windsor chair, a shaped solid-wood seat into which arc-socketed wood legs and spindles, appeared in England sometime in the late 17th century. While the Queen Anne designers were dogmatically building rigid chairs founded on their refined understanding of mortise-and- tenon joinery and the strength of materials, Windsor makers had begun drawing on an intimate that are lighttveight yet strong. Rear, from left to right: continuous-arm chair, fan-back armchair, high oval-back desk chair, birdcage sidechair. In the foreground are two stools and a child's fan-back.

knowledge of the nature of materials, much of it learned in other woodworking trades, to build comfortable, flexible chairs that could sell at half the Queen Anne price. Figure 1 on the facing page shows how these mass-producible parts work together as a team, like the parts of a suspension bridge or a bicycle wheel, branching together into a single tension/compression unit. Where the Queen Anne chair, built like a post-and-beam house, required eight or more braces (four disguised as ear pieces at the knees), the Windsor needed none. The whole chair was itself a brace.

Cherchez la femme?-It was a timely idea, and a woman may have been its inspiration. After the fall of Rome, only great men sat in chairs, or more properly, thrones—heavy, complicated and expensive. It wasn't until the 16th century that women, leading their own lives for the first time, demanded chairs of their own, specifically lightweight, portable chairs to take into their all-female bed-chamber social gatherings. Among these were the first crude precursors to the Windsor.

I sometimes wonder if they weren't collusions between the woman of the home and the one branch of the woodworking trade-that was her domain. Historians generally agree it wasn't a city chairmaker who invented the W indsor. Some suspect a turner, and otilers a wheelwright, but as one who has repaired both old Windsors and old spinning wheels, I can say that to a spinning-wheel maker, building a Windsor would have been as easy as making the flyer 10 sizes too big and shaving the wheel spokes into spin-

Pliolo Andrew tdgjr. Unwind I >.i, id IXinn

Traditional Furniture Projects 111

Michael Dunbar

Michael Dunbar

Dunbar uses traditional methods and tools to make his chairs, a low-overbead approach that produces two chairs per week and allows a comfortable profit.

dies. Everything else—the splayed, turned legs in tapered sockets beneath, the wheel posts (translated to arm posts) in similar sockets above, and even the ornamentation—is identical with the parts of Windsor chairs. In fact, a wheel's functional reels, rings and tapers translate into one of a Windsor's main production economies. All die decoration on a Windsor can be done by machine.

At any rate, wherever it may have originated, it was almost certainly near Windsor, England, that the design matured, for from the 1720s onward, it was to the southern counties, and Chipping Wycombe in particular, one went to buy these green-painted "stick" chairs that were becoming all the rage as garden seats.

Storming the colonies-With a name like "Windsor." the chair might have become synonymous with tea and crumpets, but by 1750 it was the American Windsor that was known worldwide. Given the circumstances, such a flourishing was inevitable. First, a burgeoning American population—in Philadelphia alone, from 13,000 in 1740 to 40,000 in 1776-had reached a degree of civilization that required fashionable chairs, and furniture then, as now, was not easy to ship.

Second, we had die wood. English Windsors were almost always all of hardwood, an expensive material not conducive to easy-production. After centuries of shipbuilding, they had few large-softwoods left, while we had—still have—whole forests of clear, straight pines, poplars and basswoods wide enough to make more than 50 18-in.-wide seat blanks per tree. (1 can look out and count 10 in the Vermont woods outside my window right now.)

Third, we had the technicians. In the previous century, essentials were the priority, giving rise to a massive guild of turners and spinning-wheel makers. Many of these had by now begun to slip into the niches opened by the more genteel life-style, offering rush-bottom and banister-back chairs and daybeds among their wares, then advancing to Windsors as these began arriving from abroad. At first these domestic copies were considered stopgaps for those who couldn't afford the real thing from England (George Washington dismissed them as too flimsy "for common sitting" and ordered his from England), but by 1755, things British were already becoming tainted. Washington, in fact, eventually canceled his order and furnished Mount Vernon with American Windsors.

Finally, our developing democracy was a fertile social milieu for mass production, a concept in its infancy elsewhere. The rising middle class created such a huge market for mid-price furnishings, that advertisements from the 1760s onward described batches of thousands of Windsors—often all from one shop—for sale in major ports all along the coast, the West Indies and Europe. These chairs, made by the likes of Francis Trumble and Joseph Henzey, were often branded with the maker's name-to prevent mix-ups during shipping, the beginnings of the "brand name." Virtually all of the early advertisements originated at Windsor "manufactories" in Philadelphia, the mid-century Windsor center of America. In fact, Philly enjoyed such a monopoly, that by the 1760s all American Windsors were becoming known as "Philadelphia chairs," a trend New York and New England makers found distressing.

They set out to rectify their subordinate position in the good old American way: by building a better product. By introducing crisply executed turnings, bold carvings and daring splays and proportions, the northern Windsor makers gradually weaned their neighbors from out-of-state shipments of standardized mass-produced designs. Along the way, their experiments toward a more appealing Windsor led to die development of die continuous-arm design, perhaps the boldest of them all and one apparently never made south of New York.

North of New York, today: Dunbar in Portsmouth-You can't research the history of Windsors very far without bumping into the name Michael Dunbar. Dunbar didn't just write about traditional Windsors, he built them—a lot of them. From 1970 to 1985 he built well over 1,000 chairs using the original methods and tools. Along the way, he wrote several books on the subject and became something of a Windsor guru, inspiring a subculture that now includes at least 100 Windsor makers across the country. Here was the man to see.

I met with Dunbar at the art museum at the Rhode Island School of Design (RISD). where we looked over the museum's collection of antique Windsors and talked shop. Much of what Dunbar said addressed modern misconceptions. "In the 20th century we're in love with wood, to the point where 'the natural beauty of the wood' has become a cliche. Everybody says: I want to make a Windsor chair, but I want to make it all out of walnut. But a Windsor is an engineered product-the wood has been selected for its properties, not for what it looks like." Which is why the old-timers chose only tough, resilient woods like oak, ash and hickory for their spindles. Besides having the strength to support and the flexibility to conform to the sitter, these woods rive and shave easily enough to enable a good worker to turn out a set in an hour. The legs and armposts were made of maple, beech or birch, woods that, when turned green, cut so cleanly an experienced turner could knock out a turning every three minutes. For the seats, they chose a single slab of clear white pine, basswood or poplar, wide enough to avoid the extra step of edgegluing and soft enough to saw and adze into shape easily.

These tricks work as well today as they did then, but we power toolers have become so conditioned to doing things by machine that we tend to work with machines even when it doesn't make sense, to die {joint where some of us end up reinventing the wheel.

Dunbar went on to say, "I once read an article by a fellow who began by saying: It may be more satisfying to carve out seats with an adze and inshave, but I run a production shop ' He then goes on to describe how he uses a chainsaw to rough out his seats, and his last sentence is: 'And doing it this way I can make a seat in two hours.' Hell, I can do it in half an hour using traditional tools! Which brings up a point about business. Too many people believe unless you've got the drill press, the jointer, the bandsaw, the lathe, etc., you can't be a woodworker. I make two chair seats a week using a S45 bowsaw. It takes me from five to seven minutes to cut out the two seats, after which the saw goes back on the wall 'til next Monday. A bandsaw takes up space and has to be cleaned off, and it costs a lot more than S45. It's ridiculous for two or even six chairs a week. You have to decide what level of production requires setting up machinery."

Dunbar is able to turn out two chairs a week in his New Hampshire shop because the Windsor was conceived not as a show piece but as a production piece, built of the easiest-to-work woods by the fastest methods. The unfinished product, a mix of wood colors with tool-marked surfaces, was brought together visually and made smooth with paint, which Dunbar feels adds to the challenges. "Anybody can take crotch mahogany and make it interesting. That's no real achievement; it's pretty to look at anyway. Now take a product that's going to be painted and make it look interesting. You've narrowed yourself down to a very tight range of possibilities, and one of them is line. I'm convinced that among the "mysteries' of the trade, those things handed down by oral tradition, were instructions on consciously working in the element of line."

The Windsor is a composition in which each of the parts moves the eye to the next in a continuous linear flow. Because the eye is always moving through the chair, individual parts need not be perfect. Dunbar pointed out some alarming discrepancies on RISD's old Windsors. He suggested some chairs had one or two skinny legs because of rough blanks that were rived too small from the log, or because the turner tried to shave away a slipup with the skew chisel. Discards? Those legs were made the hard way: "Once we've got it chucked up and roughed round, there's no way we're gonna throw that thing away. It either blows up on the lathe or it goes on the chair.''

You'd think the maker would at least have sorted his legs so that each chair had, say, two fat legs in the front and two skinny ones in the back, but generally in the period they seem not to have recognized that as a convention. "We've grown up with ma-chine-made uniformity, but back then if a leg had a flaw (he points out a thumb-size knot hole on a leg's front), it went where it went and there was no conscious effort to turn it to the back."

The fact is, they didn't have the time, because these guys were making them by the thousand. The profit margin was too slim to allow for finickiness and still is, as Dunbar's experience shows.

"You learn these things the hard way. For the first five or six years in business, I'd make anything anybody wanted, until I realized I was losing my shirt. Finally I got kind of hard-ass about it. I made the kid's chair, the oval-back side chair, the sack-back and the continuous-arm, and that was it. Not only did you have to come to the shop to pick up the chair, you had to come in to order it, because I didn't want someone to order a chair and then come in and say 'My god, this isn't what I wanted at all!' So you'd come in, look at the model chairs I'd work from, and that's what you'd get." Dunbar's only display was in his kitchen. His deposits were nonrefundable, he refused to ship and he sold the chairs painted or unfinished only.

Buyers weren't deterred. Collectors who couldn't afford antique Windsors paid S450 each for his chairs, and he turned them out like a one-man production line. But he kept his sanity, too. "I almost always made them in pairs. I'd bend the backs on Monday so they'd be ready to go on the chairs by Thursday. I'd turn the legs and by the time I got done with the eighth one that was about all I wanted to turn. The same with the seats. Two was enough." Was it worth it?

"It's a wonderful way to make a living "

Dunbar wasn't the first modern Windsor maker to revive the style—that honor probably goes to Wallace Nutting of a prior generation—and he wasn't the first living Windsor maker I'd ever met in person.

Sawyer in Calais—In the summer of 1982, I served on the jury at Frog Hollow Craft Gallery in Middlebury, Vt. Dave Sawyer, a tall, slim, prepossessing and affable gentleman with lantern jaw and expressive gray, bushy eyebrows had just lugged into the jury room two of his early efforts, a blue sack-back and a red youth chair. The chairs were okay, but they lacked charm. As the woodworker on the panel, my "no" vote was the clincher for Dave's rejection.

About six months later, he returned with a beautiful continu-ous-arm and one of the most beguiling and comfortable fan-backs I've ever tried—obviously a fast learner. The gallery has been showing his work proudly ever since, and Dave and I have been on close terms, visiting and following one another's work from time to time.

Sawyer graduated from die Massachusetts Institute of Technology with a mechanical engineering degree in 1959. He'd taken up the profession because, after helping his dad, a commercial artist, in his professional workshop since childhood and working in a boat yard while in high school, he knew he wanted to be involved in building things. He took a nine-to-five job as an engineering draftsman in 1963, but says: "It put me to sleep, so I gave it up. It was all pretty far removed from actually making something."

As an antidote to pencil pushing, he built an 8-ft. by 10-ft. house from scrap lumber in a week on six acres in Quaker City, N.H. Having no room for woodworking, he proceeded to make leather belts, then candles. After adding onto the house, he graduated to woodworking with a line of wooden pitchforks.

In 1970, a thousand or so pitchforks later, he built a shop. Having found a ladder-back he liked at a friend's house, he went into ladder-back chairmaking, a trade he chose because he "liked the idea of starting with a tree and a few hand tools one day and having a finished product the next."

Sawyer eventually took a trip to see Dunbar, who at the time was giving all comers measured drawings remaindered from one of his books. Dave took a set home and finally built his first Windsor, a sack-back.

Sawyer constantly experimented with improvements. He curved the back of his continuous-arm more deeply than those of antique models, making it more inviting to sit in, "more like sitting in a basket and less like sitting against a wall." A more daring experiment was to try an oak back on a butternut seat and cherry undercarriage (see the middle photo, next page). Showing through a clear oil finish, this makes a pleasing combination, and

since butternut is as easy to carve as pine, and cherry as easy to turn as the traditional white woods, it adds little to the construction time. Sawyer himself still prefers his Windsors painted, but says, "I like them both ways. A lot of customers like to look at wood, and I don't mind humoring them."

Sawyer's comb-back rocker, shown above, is truly an avant-garde chair, since there's almost no such thing as a traditional Windsor rocker, rockers being invented and popularized later. To me, it's an intriguing taste of where Windsors might have gone had they kept going. Always on the lookout for new ideas, Sawyer travels from Quebec to North Carolina to talk to his fellow craftspeople, constantly learning and often teaching about ladder-backs and Windsors. Ironically, he met one of his closest Windsor-making friends, Robert Chambers, just over the mountain in Corinth, Vt.

Chambers in Corinth—A fellow engineer, Chambers started really high-tech as an aerodynamic designer of jet missile engines. "I was making lots of bucks in aerospace but also suffering a big dose of questioning what I was doing. From the earliest times, I remember being with my father in his woodworking shop—he didn't make much furniture but was always doing something. I'd get a real craving to do something and find out right away if I did it right or not."

Chambers picked up a copy of Dunbar's old book. "I found it tremendously appealing that I could go knock some trees down one day, stand against the lathe with spray in my face the next, and in between fool with the design."

He took his first sack-back to Dunbar, who pronounced it one of the best first efforts he'd seen, but Chambers, his own harshest critic, was convinced Dunbar was just being polite. He soon got a chance to try again when a friend asked if he'd make him a pair of sack-backs and four side chairs. It was then that it occurred to Chambers that he might be able to sell these things. "I made the decision to jump out of my comfortable position and move to Vermont," says Chambers. "I took out ads in a few trade journals, and since my prices were low (SI75 to S275) and there weren't many other Windsor makers at the time, I got myself terribly busy, more than I wanted to be." He dropped the ads and cut back to take a teaching job at Vermont Technical School in nearby Randolph until, a year ago, he felt ready to regroup.

Meanwhile, his sack-back design evolved. He bent the back bow on his arm-crest form so that, with the same curve to both bends, the chair looked unified. He took the same kind of thinking further with his undercarriage layout.

Many Windsor makers, like Dunbar, "know all the angles" and simply eyeball their leg angles. Chambers prefers to stand a block with a guide mark on his seat blank to help him align the bit when he bores his leg holes, using a modern bit and brace rather than the spoon bit of old. Then, to ensure symmetry, he bores the rear leg holes using die block to set his side-view angles only. The rear legs' rear-view angles are always bored in the same planes as those of the front legs, like a rocking chair's, so they can be sighted directly from legs plugged into the front holes. Besides facilitating boring, this trick seems to serve equally well as a design formula, for the result, in Chamber's hands at least, is a chair with strikingly well-balanced proportions. All his Windsor styles, with their rear-leg tenons set well in toward the seats' center-lines, remind me of lithe dancers up on their toes.

Dave Sawyer Chairs
Pholo: Jav O'Rcar

Dave Sawyer

Dave Sawyer, left, captures the flavor of Windsor style in overall proportions as well as in crisp details. Sawyer's comb-hack rocker, above, is a style that may have never existed in the period; rocking chairs were rare. Sauyer's reproportioning of the cherry legs, necessarily short and wide-splayed, is pleasingly well-balanced. The seat is butternut and the superstructure oak. The Windsor accepts stylish detailing without sacrificing its underlying integrity of structure. In the beginning of the 19th century, tvhett things Chinese ivere in vogue, bamboo-pattern side chairs, such as the one above right, were popular. Simplified detail streamlined production.

Robert Chambers

Robert Chambers, left, here drilling tapered sockets with a reground spade bit, is a former aerospace engineer who continually refines his Windsor designs. Typical of Chambers' Windsors, shown above, is an uncluttered, unified look, much in contrast to some early chairs (example top left). Homely furniture from the past is a reminder that utilitarian forms weren't always built by design geniuses.

"Each time I go to make a chair, it's an opportunity to make it better. While some of my chairs are pretty close to standardized, most are variations on a theme.

"1 accept a lot of the given," concludes Chambers. "I haven't invented any new styles, but I still find it a big challenge to put together something that works, that's pleasing, and harmonious with an earlier time."

Over all. it takes Chambers about 40 hours to build a continuous-arm from tree to paint. When I hinted this was a mite longer than the old-timers took, he replied. "I've hewed one side of a 26-in. 8x8 in an hour and a half, so I can see where it could be done in the six hours it was supposed to have taken 'those old-timers'—if you could keep it up. But, you know, a friend of mine-told me once, 'Hey, Bob, I've seen an old photograph of some of those old-timers. They're about 26."' Chambers is a bit more mature than that.

Both Sawyer and Chambers have restored "bodgering"—turning Windsor parts-to the fine art it once was. But to my eye, there's a man who has taken a step beyond.

Franklin In Easthampton-Born in Burlington, Vt., one of Peter Franklin's earliest memories is buying his first antique while accompanying his parents to an auction as a small boy. When he-grew up, his interest in things old led him to a career as a contractor in Nova Scotia, where he restored buildings. But one winter day, standing on a half-finished roof in a freezing rain, he found himself asking, "Why am I doing this to myself?" A friend suggested an apprenticeship with a man who made Windsors in Bristol, R.I. Franklin made the move.

As an apprentice, Franklin made thousands of parts, but no chairs. Since there was a master turner in the shop, producing a leg every three minutes, Franklin was left to figure out the lathe-on his own, after hours. After leaving his apprenticeship and moving to his present shop in Massachusetts, he continued to fine-tune the skill, beginning with a reproduction of a New England-style leg he'd measured from a chair belonging to a couple in Westchester, N.Y. He learned well. His turnings are so perfect it's hard to believe he hand-turns these patterns by eye, again and again, for every chair he builds.

His shapes are crisp, delicate, bold and daring all at once, but it's a feat he had to work a compromise to achieve. Franklin turns kiln-dried maple, which he chose for its great strength and density. "These turnings can be light in form yet retain great strength. Additionally, maple's density permits a sharpness of detail not possible in softer woods," says Franklin. By turning the wood dry, he avoids the crumbling and fuzzing of green stock, enabling him to turn clean curves with razor-fine rings and shoulders. He admits it's a trade-off, as it takes considerably longer to turn dry-stock. particularly maple, than it does to turn traditional woods.

Franklin spent the first few years after his apprenticeship collecting designs. One of his best is a variation on a continuous-arm that had come into the Leeds Design Workshops, in the same building as his shop on 1 Cottage Street, for repair. After acquiring permission from the owner, he applied his standard routine. "I take a tape measure and paper and draw little lines corresponding to significant features. Then I walk away from the chair." The result is an interpretation, not an exact duplicate.

In the shop. Franklin tranfers die chair's seat outline to particle-

Photos above and below right: Peter Franklin

Peter Franklin

Peter Franklin, above left, who works with a couple of helpers, combines high production with snappy detailing that's almost audible. Franklin's settee, above right, is a Windsor form still evolving: The vibrant undercarriage appears to be impatiently waiting for the sedate top to catch up. Franklin's comer chair, at right, is a form that probably never existed, but it dearly carries on the tradition toward ever more buoyant design. The earliest Windsor precursors, drawing below, a handful of which survive from the 16th century, were mired in Gothic heaviness. Lightiveight intermediate forms, photo below, realized the possibilities of tension/compression engineering.

Tension Engineering

Museum of An. RIxkIc Island School of Design: gift of the estate of Mr- (iustav Radckc

Photos above and below right: Peter Franklin

Museum of An. RIxkIc Island School of Design: gift of the estate of Mr- (iustav Radckc board and cuts it out on the handsaw. Then he glues hardwood blocks at the leg locations and uses the leg angles he's recorded to bore rough guide holes. This unit becomes a combination seat pattern and boring jig. Or, if the seat shape is one he already uses, he may grab a finished blank from his stock, glue on the blocks and bore it directly so he can stick in whatever legs he was kicking around and see how it looks. It doesn't always look right, but Franklin is philosophical: "You can't be paralyzed by fear of ruining a good seat. You've got to take a chance, hold your breath and drill it."

Even if an experiment comes out looking good, it still may not be right, because for Franklin, the chief determinate is comfort. "The philosophy of the early makers was to make chairs to sit in. They weren't making art furniture or chairs to set aside in collec tions. and if it's not comfortable, you can't use it."

Nor were they pedagogues. Thus Franklin feels invention is justified. "I just built something that may never have been made before, a Windsor corner chair, for a show in Holyoke, Mass. If the original makers had had access to as many things as we do, I think they'd have been as whimsical as we. Just because you're working in traditional elements doesn't mean you can't incorporate new ideas into traditional design."

To my eye, Windsor design is not only flourishing, it's healthier than ever—a 200-year-old tradition that's just now getting its second wind.

feremy Singley is an author and a full-time woodworker in Fast Middlebwy, Vt„ where he makes Windsors and other furniture.

Green Chair Makers Burlington

Fig. 1: Evolution of period doors

1650-1720 First Period featheredge with batten

Fig. 1: Evolution of period doors

1720-1790 Georgian six-panel (High Style)

1650-1720 First Period featheredge with batten

1720-1790 Georgian six-panel (High Style)

1820-1860 Greek Revival with flat or raised panel

1780-1820 Federal six-panel

1820-1860 Greek Revival with flat or raised panel

Was this article helpful?

0 0
Wood Working 101

Wood Working 101

Have you ever wanted to begin woodworking at home? Woodworking can be a fun, yet dangerous experience if not performed properly. In The Art of Woodworking Beginners Guide, we will show you how to choose everything from saws to hand tools and how to use them properly to avoid ending up in the ER.

Get My Free Ebook

Post a comment