Four-Room House

FOUR-ROOM HOUSE: Figure 1. House plans. (a) Four typical four-room houses, and one slightly atypical three-room house, from Tell el-Far῾ah (North) (Chambon, 1984); (b) three typical and one “squeezed in” three-room house from Tell en-Naṣbeh (McCown, 1947, survey map); (c) four-room house with added side-hall or shed from Shechem (Wright, 1965, fig. 76); (d) a proposed second-floor plan of the house figured in (c) above. (Courtesy J. S. Holladay)

view larger image

Together with the three-roomed variation, the four-room house was the dominant house form in ancient Israel (see figures 1a and 1b respectively). A shed was sometimes added along one side (see figure 1c). An uncommon two-roomed variant lacked the characteristic transverse rear room. For functional and structural reasons the form was highly standardized. As a result, unusual or extensively elaborated plans should be closely examined to ensure that they do not represent architectural reconstructions of more than one building or stratigraphic phase. This distinctive house is found throughout Israel and Judah, from the earliest hill settlements (Marquet-Krause, 1949; Callaway, 1983; Stager, 1985) to the latest Iron II towns. [See Judah.] While not the only house form seen in Israelite cities, it is found nearly universally in lightly fortified desert farms (Cohen, 1979), farming hamlets and villages, towns, and military outposts.

Door placement and other details apart, the form was highly stylized almost from the start (e.g., at Khirbet Radannah and Ai), although it could be fitted into an existing space by tapering its overall shape (or by repositioning the rear room or even eliminating it) if alternative storage space were available nearby. [See Radannah; Ai.] Ethnographic parallels and architectural analysis support the conclusion that this house was highly adapted to normative Israelite family structure, modes of production (mixed peasant dry-farming and/or horticulture and animal husbandry: Hopkins, 1985) and socioeconomic organization (Holladay, 1995, pp. 391–393). The sporadic earlier appearance of similar houses in Late Bronze Age Syria is of interest and may have something to say about the emergence of early Israel.

Palestinian House Forms and Ethnicity.

In Palestine, with the exception of a few examples at tenth-century BCE Philistine sites (e.g., see Maisler [Mazar], 1951), all occurrences of the four-room house known to date seem to be limited to Israelite or Judean sites. Typical Philistine houses were quite different (Dothan, 1971; Gitin, 1989). [See Philistines.] The form also appears in Transjordan, with other styles of pillared house construction, but not enough is known of its dating or archaeological contexts there to allow for a clear interpretation. In general, however, even when the form appears at a non-Israelite site, it does so during such periods of overwhelming Israelite hegemony as the United Monarchy. In those cases, the buildings may well have housed ethnic Israelites. Though data are sparse, the form does not seem to have survived the Babylonian Exile. For all these reasons, it seems legitimate to call this the Israelite house.

Architectural Reconstruction and Functional Interpretation.

Extremely few four-room houses are preserved to the second story, most being known only from their stone foundations and floor surfaces. Thus, details of the upper story or stories, roof, and so forth must either be extrapolated from a few known details or inferred from constructional features (see below). Similarly, functional interpretations must proceed on the basis of inference from architectural details in the light of ancient and ethnographic parallels, along with a smattering of meaningful artifact distributions. As a negative case in point, rows of storejars in the stabling areas of houses destroyed by enemy action may reasonably be construed as exceptional, reflecting preparations for siege and not normal storage patterns (cf. Holladay, 1986, p. 160). As is often the case in archaeology, even the name four-room house is misleading, describing only the main structural divisions of the ground floor—one or two side aisles, a misnamed “courtyard,” and a rear room in the usual terminology—and ignoring both the characteristic second story (the structural reason for the massively pillared construction) and the frequent divisions in the side and back “rooms.” In fact, apart from their very unitary character (each household clearly is an independent social unit), the pillars and cobbled side aisles are the chief distinguishing characteristics of these houses. The pillars are of a variety of locally available materials—from well-shaped caprock (Ar., nari) monoliths to stacked, roughly dressed stone drums or slabs to heavy wooden or mud-brick pillars. Occasionally, the beams borne by the pillars were also stone monoliths or were complexly built up of cantilevered stone slabs. Only high loading factors would ensure the integrity of the stacked stone-boulder or stone-slab construction; the massiveness of all the other forms similarly attests to the tremendous load-carrying capacity of the line or lines of pillars.

Floor surfaces range from hardened surfaces composed of cobbles or flagstones bedded in ashy soil for the pillared side aisle/s to relatively clean packed earth, often multilaminated, with a few flat-lying sherds and occasional slivers of bone, for the rear room, small side rooms, and the center room or courtyard. Rarely, the center room was whitewashed or surfaced with a thin white marly “plaster.” Some small side rooms lack clear surfaces and may be interpreted as mirey, uncleaned stalls for the penned-up “fatted calf” (1 Sm. 28:24; Am. 6:4; see below). Ethnographically, the first five floor surfaces noted here are associated with such different functions such as stabling (Holladay, 1986), store-rooms and kitchens (the latter having an ashy floor), living rooms, and nonhardened (lacking cobbles or flagstones) domestic stables (Kramer, 1979, pp. 147–149; Watson, 1979, p. 121). That the so-called center room or “courtyard” is an interior surface is easily tested by an objective (e.g., granulometric) comparison of its surfaces with other surfaces known to be exterior (e.g., streets and other open areas: Kramer, 1979, pp. 148–149; Watson, 1979, pp. 157–159, pls. 5.3–5), as well as by the fact that, although the “courtyard” surfaces are often lower than exterior street levels, with high thresholds to keep water out, no provision was made for the internal drainage required if they were open to the sky.

Stairways are commonly noted and can be reconstructed in other instances. From everything presently known, the first story was low, averaging under 2 m, with considerably less clearance under the beams. It was, thus, unsuited to human habitation.

Interior Features.

Most houses had stub walls between the supporting pillars with, where preserved, mud-and-stone mangers on top. The mangers are similar to those found in stables at Hazor and Beersheba and would have been used for donkeys and, probably, cattle (the standings are too short for horses). [See Hazor, Beersheba, Stables; Cattle and Oxen.] Semicircular ground-level mangers (“bins”) are common and may have been used for sheep and goats, but possibly also for cattle. [See Sheep and Goats.] Fire pits or hearths in the courtyard are characteristic at Radannah and at Tel Masos, but in general they are rarely found. [See Masos, Tel.] Other types of hearths are found occasionally, mostly in the central space, or “courtyard” (e.g., the “pi”-shaped hearths in the Drainpipe House at Ta῾anach and in house 1727 at Shechem, and in the hearths contained in the overturned upper portions of storejars sunk into the ground in buildings at Shechem, Hazor, and Meṣad Ḥashavyahu). A large industrial hearth, which replaces an earlier wine press at Shechem in house 1727 seems to be unique. [See Ta῾anach; Shechem; Meṣad Ḥashavyahu.]

Hearths are by no means commonly found. However, it is clear that every household must not only have had cooking facilities (cooking pots are one of the most common vessel types), but, in most parts of the land, also hearths for heat in winter. (Ovens served for bread making, not cooking.) Ethnographically, hearths are one of the chief markers of living rooms (Watson, 1979, p. 122, fig. 5.2; Kramer, 1979, pp. 147–148). These are rooms within which the entire nuclear family cooks, entertains, eats, sleeps, and performs daily indoor activities. Families having two living rooms use the second one for entertaining or rent it. At Patty Jo Watson's “Hasanabad” and “Ain Ali” bread ovens were not used: one hearth served for winter warmth, cooking, and—through the use of the saj, the modern descendent of the baking tray—baking (Watson, 1979, pp. 205, 283). During the summer, cooking fires were relocated into exterior courtyard spaces. Carol Kramer (1979, pp. 147–148, emphasis added) presents illuminating data: “Hearths are peculiar to living rooms.…In contrast to ovens, such hearths, found throughout the Zagros…and elsewhere in Southwest Asia, are often situated in second-story rooms, and as such have been found at an archaeological site near Shāhābād.” Ovens, on the other hand, are generally placed out of the wind at points of easy access from outside (e.g., Tell el-Far῾ah [North]; Chambon, 1984); they are not normally found in or adjacent to every house. This implies communal use, as in present-day traditional villages. [See Far῾ah, Tell el- (North).]

Silos or grain pits have been found in the rear room and elsewhere (Watson, 1979, pp. 125–126, fig. 5.4). Multiple exterior grain pits were common before and during the United Monarchy but are not characteristic of later periods (Holladay, 1995, pp. 377–379, table 3). Vermin-resistant above-grade grain and flour bins or chests constructed of mud and chaff are common in ethnographically witnessed villages in Iran (Kramer, 1979, pp. 144, 147; Watson, 1979, pp. 67, 162, figs. 5.42–43), suggesting that they might have been used for domestic storage in Israel; none have yet been identified, however. Where cisterns were used, they would have been located either beneath the house or outside. [See Cisterns.] In some cases, houses may have been built over preexisting cisterns.

Functional Aspects of the Ground-Floor Plan.

The ground floor in the four-room house falls entirely within the family's economic domain. Stalls for donkeys and large cattle are readily identified by their similarity to standings in chariotry stables (cobbled or flagstoned surfaces, bench-style mangers, pillared construction). Floor-level mangers may attest to the feeding of sheep and goats. By analogy with modern European modes of veal production, small rooms with blocked doorways and miry surfaces may have housed the “fatted calf.” Small storerooms with packed-earth or mud-plaster surfaces probably served for specialized storage, while the back room seems to have been the main locus for heavy storage: ethnographically, the items stored in such rooms change with the season. In ancient Israel they would have included young animals, bulk quantities of wine and oil, fodder, grains and legumes, reserved seed grain, straw, dung, twigs and dung cakes (the primary fuel source), pottery vessels not in use, timbers, raw materials for craft production, and a plow, yoke, and other farm implements (DeBoer and Lathrup, 1979, pp. 110–121; Hall, McBride, and Riddell, 1973; Jacobs, 1979, p. 179; Kramer, 1979, pp. 144, 147, 154; Watson, 1979, pp. 160, 294–296). Light items—dried fruits and dairy products, spices, textiles, and other domestic goods—would have been stored on the second floor (Kramer, 1982, p. 105). The large central space seems to have functioned by day as a work area and is where large limestone mortars (used with long wooden pounders for pearling barley?) were typically located. At night and during winter storms this area would have served as a secure folding area for the family's small sheep and goat herd (Hopkins, 1985, pp. 245–250).

It has often been suggested that the rear room served as the family's living room, and that the Israelite house evolved from this original nucleus, perhaps originally imitating a nomad's tent (Fritz and Kempinski, 1983, pp. 31–34; Herzog, 1984, pp. 75–77). However, there is absolutely no evidence for this type's existence in the second millennium BCE. Against this, the mean width of ninety-five identifiable rear rooms in Frank Braemer's catalog (1982, pp. 160–269) is 1.98 m, while the mean widths of ethnographically described living rooms range from 2.78 m at “Hasanabad” to 2.9 m at Aşvan to 3.02 m at “Aliabad” (Watson, 1979; Hall, McBride, and Riddell, 1973; Kramer, 1982)—their width being determined by the length of the available roofing timbers. In short, a family of four to five persons cannot live, work, cook, eat, entertain, and sleep in a 2-meter-wide hallway with 1.52 m of head room.

FOUR-ROOM HOUSE: Figure 2. Selected ethnographically documented peasant housing, (a) One-story house from Hasanabad, Iran (Watson, 1979); (b) two-story house from Aliabad, Iran (Kramer, 1982, fig. 4.7); (c) two-story house from Aşvan, Turkey (Hall et al., 1973, fig. 4). (Courtesy J. S. Holladay)

view larger image

FOUR-ROOM HOUSE: Figure 3. Reconstruction of the house in figure 1: c–d, featuring a hypothetical ῾ălîyya, or upper room. (Drawing by Megan Williams; courtesy J. S. Holladay)

view larger image

Second Story.

A second story must be inferred for the four-room house on the basis of its constructional characteristics: its massive load-bearing pillars in a roofed-over courtyard and lack of provision on the ground floor for living quarters or, in most cases, for cooking or heating facilities. Given the lack of these facilities on the ground floor, the second floor should constitute the family's “living domain”: the living rooms should have widths and overall floor areas similar to those in the ethnographic literature—or an average living room size of between 15 and 20 sq m (Kramer, 1979; Watson, 1979), although larger and smaller ones do occur. Living rooms should have other dimensions (e.g., headroom), facilities (e.g., hearths), and functionality not available on the ground floor (cf. Hopkins, 1985, pp. 144–148). An entry hall at the head of the stairs, used to store outer clothes and men's portable tools, is suggested by the ethnographic data. Kitchen facilities and one or two light storage rooms would round out the second floor plan (see figures 2b, 3). The second story's walls may have been thinner because of their reduced load-carrying requirement. This would have increased its area slightly. A third-floor ῾ălîyyâ, “upper room, or roof chamber,” preferably accessed by an external staircase, might also be present, whether for rent or use as a guest room (see figure 3). Balustraded windows, unsuitable for ground-floor use, would be perfectly in place in the higher stories, providing good light and ventilation. Domestic versions would have been more rustic than the palatial one featured in the “woman at the window” ivories from Nimrud. [See Nimrud.]

Roofs were flat, of packed mud and chaff (“straw”), with new layers applied annually. Following a rainstorm they were rolled with limestone rollers or carefully compacted by foot (Wright, 1965, p. 161, fig. 80; Watson, 1979, pp. 119–120). Sprouted grain on the rooftops was a proverbial sign of spring (Ps. 129:6; Is. 37:27). Parapets, stipulated by Deuteronomic legislation (Dt. 22:8), were necessary to enable the use of the roof as an outdoor workspace (i.e., it would have functioned analogously to enclosed courtyard space) and as a safe place for women and their small children to socialize. With no roof overhang, rainwater runoff would have eroded the mud-brick (adobe) fabric of the house. Thus, jutting wooden (or ceramic) “gargoyles” (Watson, 1979, p. 159) or internal drainpipes leading to an in-house cistern (Lapp, 1967, pp. 2–39, fig. 14), would have been needed to drain the roof. The gargoyles' individual splash areas or runoff channels (Watson, 1979, p. 159) would have been visible on exterior, street, or lane surfaces. Ethnographic parallels show that brush shelters (biblical sûkkôt, Neh. 8:16) are set up to soften the summer sun and that during the hottest months, families often sleep on the roof. Ethnographically, this is the woman's domain—house tops being the common pathways between closely packed neighboring households in some modern walled villages (Jacobs, 1979, p. 179). There, and by parallel in ancient Near Eastern cities and villages, cultic ceremonies are conducted on rooftops (Jer. 19:13; Zep. 1:5), as were corporate mourning and rejoicing (Is. 15:3, 22:1), sleeping (1 Sm. 9:25), bathing (2 Sm. 11:2), washing, and clothes drying, basket making, baking, and drying food and fodder for storage, dung and brush for fuel, and flax and other agricultural products for processing (Jos. 2:6). They probably were also used for storing beams and large household and farming implements during the rainless months (Jacobs, 1979; Kramer, 1982, p. 111; cf. Hall, McBride, and Riddell, 1973, p. 248).

Economic Significance.

Except for the large numbers of grain silos (averaging a storage capacity of about a ton each) occasionally found near certain houses prior to and during the Solomonic era (Greenberg, 1987; Finkelstein, 1986, pp. 1–20, fig. 4), but not met with later, large-scale facilities for accumulated agricultural surpluses are not known from the period of the Israelite monarchy. This stands in marked contrast to the military and economic policies of the Persian period, as witnessed by the large number of great store pits at such sites as Tell el-Ḥesi (Bennet and Blakely, 1989, p. 67) and Beersheba (Naveh, 1973). [See Ḥesi, Tell el-.] Nor are redistributive mechanisms, such as the large-scale bakeries found in conjunction with granaries in Saite through Ptolemaic Egypt (at Tell el-Maskhuta) known from either Israel or Judah. Even the three hundred-odd jars filled with oil, wine, and grain in the Phoenician “fort” excavated by Zvi Gal (1992) at Ḥorvat Rosh Zayit fall into different focus when it is realized that it would have taken at least ninety-one jars to store the subsistence and seed-grain requirements of one typical Israelite household and its flocks and herd, to say nothing of their olive oil and wine. In terms of scale, the storage capacity of the back rooms of only thirty-one typical Israelite houses more than equaled the capacity of silo 1414 at Megiddo, the largest storage facility presently known for either Israel or Judah (Holladay, 1995, p. 393).

Except for chariotry, frontier forts such as Tell Arad, palaces, and—perhaps—the cultus, neither ancient Israel nor Judah seems to have operated on a redistributive basis: there could have been no widespread issuing of grain and other rations to large numbers of the civilian population. Simply put, Israelite peasant farmers operated with a house (and agricultural regime) specifically designed to accumulate surpluses. The specific mechanisms by which those surpluses entered either domestic or international commerce are matters for future research, although it should be noted that a whole series of small ports, conveniently spaced about a half-day's journey apart, existed along the Palestinian coast during the Iron II period, providing a cost-effective mechanism for the long-range movement of such low-value, high-bulk goods.

[See also Building Materials and Techniques, article on Materials and Techniques of the Bronze and Iron Ages; Ethnoarchaeology; and House, article on Syro-Palestinian Houses.]

John S. Holladay, Jr.