Dr. Snelling argues that although preserved footprints provide an apparent challenge to Flood geology, namely that dinosaurs could not have left such prints in the course of a global catastrophe, the evidence from this case is not only consistent with the Flood model but contradicts conventional geological interpretations. In this article, I will look more closely at the conventional interpretation regarding these footprints and the rocks in which they are found, in order to analyze the strength of his challenge. Before I do this, however, let’s take a look at his argument.
A proposed dilemma
Dr. Snelling begins by noting challenges on both sides of the argument: “with the Flood waters covering the entire earth, the dinosaurs would have nowhere to walk. Even if they did, the churning waters would erode away any footprints left behind…on the other hand…if geologic change takes place slowly, surely footprints made in mud would be obliterated by wind and rain long before the prints were covered by new sediments and hardened into rock.” We’ll explore these challenges later.
It is not Dr. Snelling’s intention, however, to argue the mechanics behind the preservation of footprints. The bulk of his argument is geared toward the nature of this particular case, since the tracks were formed in dolomite. He asserts that dolomite is formed either locally in extreme environments (not suitable to dinosaurs) or regionally in hypersaline marine or lacustrine settings (like the Persian Gulf or the Dead Sea; also not suitable to dinosaurs). Thus it should surprise us to find dinosaur prints in such a peculiar rock.
Unless, of course, we allow for the possibility that chemically distinct marine sediments were deposited catastrophically in waves during the Flood. In this scenario, dinosaurs (and other creatures) would be threatened for their lives, and these tracks would rather reflect their attempted escape route during intermittent calm periods. Lime sediments were carried in from shallow marine waters and then exposed when the water receded between depositional events. Catastrophic plate tectonics are cited as the mechanism behind these tsunami-like events, so we might expect that volcanism played a role in chemical alteration of the lime sediment. In fact, Cretaceous sediments in Israel are interbedded with volcanic tuff layers and localized lava flows (Segev et al., 2002; Segev, 2009), which Dr. Snelling could cite as support. The importance of volcanism, he says, is that it would elevate temperatures and add magnesium to the carbonate-saturated waters, producing large quantities of dolomite.
On the surface, this argument sounds plausible. If the rock type suggests a marine environment but the fossil evidence (dino prints) suggests otherwise, there is a direct contradiction in the conventional interpretation. Furthermore, he offers a mechanism by which dolomite was laid down, and the model seems to explain the relevant data. Unfortunately, proposing a ‘plausible’ hypothesis is only the first step of scientific investigation, and there are easy ways to test the underlying assumptions here.
Back to the basics: What exactly is dolomite? And how does this relate to dinosaurs?
In case you’re not familiar with the mineral dolomite, let’s take a closer look. Carbonate rocks (limestone, dolostone) are those composed primarily of minerals that contain a carbonate ion (CO3). In pure calcite or aragonite, the carbonate ion is bonded to calcium (CaCO3), while in pure dolomite, the carbonate ion is bonded to an equal mixture of calcium and magnesium ([Ca,Mg]CO3). Since dolomite does not precipitate in ‘normal’ marine conditions (i.e. average temperature and salinity of the world’s oceans), marine sediments contain mostly calcium carbonate (CaCO3).
Initially, this seems to present a problem. Massive sequences of both limestone and dolostone (sometimes miles thick) are present throughout the world. But if dolomite does not form in normal ocean conditions, where do thick dolomite bodies come from?
It is true that dolomite forms in some ‘extreme’ environments, such as hypersaline lagoons and lakes. It is not hypothesized that large dolomite bodies formed in “oceans with unusual chemistry”, as Dr. Snelling proposes in his article, but rather that restricted circulation to the open ocean (e.g. Persian Gulf) can raise the salinity or concentration of magnesium (both of which promote dolomite formation). Most dolomite is diagenetic, however, meaning that it was originally limestone that was later modified during burial. Chemical alteration can occur through the interaction of freshwater and marine water, or the mobilization of cations (like Mg2+) from clay minerals in adjacent layers (Brigaud et al., 2009).
Two additional processes are 1) hydrothermal alteration from deep, hot fluids that moved through faults and fissures (e.g. Tritlla et al., 2001; Sha et al., 2010), and 2) microbial mediation of cations (e.g. Sadooni et al., 2010). In the former example, dolomite forms locally, cuts across sedimentary layers, and forms unique ratios of oxygen isotopes. Therefore, it is very easy to identify hydrothermal dolomites using field and laboratory analyses. In the case of microbial mediation, bacteria living in anoxic pore spaces of sediments produce excess magnesium during the reduction of sulfate (i.e. magnesium is a waste product during metabolism by certain bacteria). This process results in the regional formation of dolomite without invoking “unusual chemistry” in the oceans, and it is particularly effective in intertidal (the zone between low and high tide) and supratidal (above high tide) environments (Sadooni et al., 2010).
A minor detour
How do you know whether carbonate rocks formed slowly in ancient oceans or rapidly in churning seas during the Flood?
One common misconception is that limestone rock is simply composed of calcite crystals, while dolostone is a rock composed of dolomite crystals. Limestone can be broken down into dozens of categories, however, based on the abundance, type, and origin of grains and mud. Grains can include anything from shells, microfauna (tiny shells), carbonate sand (like on a Bahaman beach), fragments of older rocks, algal-bound or fecal-bound spheroids (called pisoids and peloids), coral, or even strands of calcareous algae.
Calcium carbonate mud tends to fill in the gaps, but other minerals can be present as well: sulfates, clays, quartz, and more. Sedimentary and biogenic structures are also common in limestone. These include cross-bedding, mudcracks, and microbial matting (in planar laminae or as in stromatolites), to name a few.
Taken together, these characteristics give abundant information about the environment and energy of deposition — how deep the water was, how fast it was moving, and its chemistry. One can hypothesize that if the minerals in limestone formed quickly in hot, volcanic-infused oceans during the Flood, we should find certain chemical characteristics. If the lime sediment were deposited rapidly in muddy waves during the Flood, we should find certain physical characteristics. These characteristics are not found in fossil-rich limestones of the Earth’s crust, however, and so Flood geology fails both tests (see next section). Thus by walking up a hillside composed of limestone layers, one can retrace the history of changing environmental conditions as the sediments were being deposited in ancient oceans.
If you’ve already become bored at the thought of interpreting carbonate rocks, then I would like to reassure you that you are not alone. Many geologists share a general disdain for carbonates: they’re confusing and tend to tear holes in your clothes during field trips. At the same time, the complexity of carbonate rocks allows us to understand numerous geological processes all at once. So pressing on, let’s consider the relationship of dolostone to dinosaur tracks.
Stratigraphic dolomites in the rock record: case of the carbonate platform in Israel
The tracks reported by Avnimelech (1962; 1966) are located in the Soreq Formation (Sass and Bein, 1982), which is part of the Cretaceous Judea Group. While the Judea Group consists largely of thick limestone and dolomite layers, the specific kind of limestone and dolomite varies geographically. In other words, if you trace the layer of “dino”-dolomite to the northwest, you will find that it transitions to fine-grained limestone and dolostone typical of a lagoon setting (in the central Israel region), then into coarse-grained limestone containing abundant rudist corals (in the Carmel region), then back into fine-grained limestone with broken shell and coral fragments (shelf break) and finally into shale (continental slope) (Buchbinder et al., 2000; Bachmann and Hirsch, 2006). In other words, there is a logical order to the interpreted environments, which represent deposition on a carbonate platform (Sass and Bein, 1982). A similar transition could be seen if you started on a beach in northeastern Australia and travelled northeast across the Great Barrier Reef.
Dr. Snelling’s assertion that the lateral extent of limestone and dolostone implies that the “Judea Group was probably formed in a vast ocean sitting over the entire region” is somewhat misleading. His oversimplification overlooks the fact that a majority of carbonate rocks in Israel formed in very shallow water, and that many were frequently exposed to the air (particularly those rocks near modern Jerusalem). Thus it is nonsense to rule out the possibility that dinosaurs (or any other terrestrial creature) could be living in the area and leave footprints. A geographical reconstruction of the region, using interpreted depositional environments, suggests that during the Cretaceous period, much of western Israel was covered by shallow seawater that was semi-restricted from the open ocean by rudist coral reefs to the west. The shoreline ran north-south, approximately between Galilee and Jerusalem, but migrated to the east and west many times in the Cretaceous (Buchbinder et al., 2000).
Sass and Katz (1982) explored the origin of dolomites comprising the Soreq Formation, and tested various models using geochemical data. Their findings suggest that the dinosaur-bearing dolomite is diagenetic, in which Mg replaced Ca and Sr in existing calcite sediments during burial. They also ruled out the possibility that it was formed during intense evaporation in an arid environment (i.e. the modern Dead Sea or Arabian shore), undermining Dr. Snelling’s claim that “the best explanation [conventional geologists] can suggest is that, for some reason, a dinosaur walked across an intertidal mudflat in an arid region (where there was nothing for him to eat!).” On the contrary, numerous dinosaur fossils are found in coastal settings. I’ve personally recovered many (theropod teeth in particular) from the western United States, where sediments accumulated along the Cretaceous Interior Seaway nearly 100 million years ago.
Finally, there is no reason to believe that dinosaurs (theropods in particular) would be confined to humid, tropical settings. Modern reptiles are commonly the most successful fauna in hot, dry climates. Nonetheless, clay mineralogical analyses by Gertsch et al. (2010) suggest alternating humid and semi-arid conditions in the Mediterranean region during the Cenomanian (mid-Cretaceous), precluding the notion of a ‘hermit’ theropod.
Gratuitous assertions vs. tested hypotheses: a tale of two models
I mentioned earlier that Dr. Snelling’s proposed model seemed plausible at first, since it contained a consistent explanation of relevant data, but that the model was easy to test. The reason is that much of the data needed is already available in previously published studies. Here is my assessment.
In Flood geology, catastrophic plate tectonics would provide the mechanism for sediment transport and deposition over Israel. In other words, massive earthquakes and shifting plates would drive tsunami-like currents over the continents. However, these carbonate rocks are not a disorderly mixture of lime mud, shells, and more, but form regular (cyclic) sequences in a logical order that resembles a range of modern depositional environments (Sass and Bein, 1982). For example, some layers contain bedding consistent with nearshore wave activity, while others contain no bedding (quiet water) or even mudcracks. Lenses of shale, chert, phosphorite, anhydrite, and quartz geodes can be found, which only form in calm waters or periods of high evaporation. These evidences flatly contradict what one might predict from a global catastrophe, but they are perfectly consistent with a model of slow deposition in a carbonate platform. Young-Earth Creationists typically respond that the apparent order of environments simply reflects repeated transgressions over the continent during the flood, in which case we ought to consider the stratigraphy in detail.
If these rocks were laid down as sediments were repeatedly washed over the continent, what would be the expected geographic distribution of rock types (think of coloring a map according to types of limestone/dolostone)? In this model, there is no reason to expect only fine-grained shale and carbonates in the outer shelf (interbedded with chalk), coarse-grained carbonates and large-scale coral reefs in the middle shelf, and fine-grained carbonates and dolomite in the inner shelf. One would rather expect a smooth transition from coarse to fine, fine to coarse, corresponding to the energy of waves. The distribution of Cretaceous limestone and dolostone can be logically interpreted in the context of slow deposition along a shallow carbonate shelf (Sass and Bein, 1982; Lipson-Benitah et al., 1997; Buchbinder et al., 2000), but simply makes no sense in terms of catastrophic deposition.
Even assuming the possibility that the Flood model can explain the distribution of sediments here, one may still consider the sheer thickness of units. Segev (2009) reports a thickness of ~1,800 meters for Cretaceous and younger carbonate rocks in Israel. Note this does not include the vast thickness of rocks underlying these units, but still requires an average of ~5 meters per day deposition over the course of a year-long flood (or a more reasonable estimate of 10 meters per day during the advance of the flood). At these sedimentation rates, it is simply not possible to form the many sedimentary and biogenic structures (small-scale cross bedding, evaporite lenses, microbial stabilization of thin laminae) seen throughout the section.
Though I wish to save the details of fossil correlation to another article, it is worth pointing out that carbonate rocks in this region can be correlated over long distances by species of microfauna — namely, foraminifera and calcareous algae. These fossils are extremely small, only visible under a microscope, and their ordered succession can not be explained by hydrodynamic sorting, potential to escape danger, or original environment (these organisms simply float around in the surface ocean). How is it, then, that the same order of species can be found in southern Israel that can be found in northern Israel (Lipson-Benitah et al., 1997) that can be found in Morocco (Gertsch et al., 2010)? Again, this is consistent with conventional models of slow deposition over a carbonate platform, but can not be explained by rapid, catastrophic deposition.
Consider also that layers of volcanic tuff are present throughout carbonate sequences in Israel. These volcanic rocks have been dated using K-Ar and 40Ar/39Ar methods (Segev et al., 2002; Segev, 2009), yielding internally consistent and concordant ages between 140 and 82 million years (the expected range, based solely on biostratigraphy). This means that the results are reproducible and that ages become progressively younger toward the top. Regardless of whether you accept these ages, it is difficult to explain why stratigraphic layers correlated on species of microfauna also yield similar radiometric ages, outside of the conventional model. But wait, there is more!
The ratio of stable isotopes from elements like carbon, oxygen and strontium in carbonate rocks can be used as proxies for seawater chemistry at the time of deposition (e.g. Saltzman et al., 1998). Thus significant changes in these ratios over time are interpreted to represent major oceanographic events in Earth history (e.g. Kump and Arthur, 1999). One such event occured in the Cenomanian, associated with the Oceanic Anoxic Event 2 (Ando et al., 2009), and is recorded in carbonate rocks from the Mediterranean region (Gertsch et al., 2010). Why is this important? Stratigraphic layers of carbonate rocks that are correlated based on index fossils and radiometric dates also contain similar trends in carbon and strontium-isotope ratios.
How does one explain this phenomenon in the Flood model? Isotopic ratios should reflect the sediment source (i.e. the chemistry of the ocean during deposition of the original sediments) or the process of diagenesis (chemical alteration after burial), but these overall trends are independent of lithology (i.e. they do not vary with rock type) and degree of diagenesis (e.g. limestone vs. dolostone). In other words, there is no reason to expect a positive spike in carbon isotopes to be present in one kind of limestone from northern Israel, dolostone from central Israel, another kind of limestone from Morocco, and limestone from the bottom of the Pacific Ocean, unless we interpret their deposition in the conventional framework: these rocks were deposited slowly in their respective depositional environments, and the isotope ratios reflect the global seawater chemistry at that time. On the contrary, the Flood model would predict a relatively homogenous distribution, or a strong correlation to rock type (reflecting the sediment source).
On a final note, Dr. Snelling proposes that magnesium and hot water added from submarine volcanism during the flood may have promoted the deposition of dolomite. However, we have already seen that dolomites in this region are geographically confined to the most inland part of the section (i.e. furthest from submarine volcanism). Furthermore, although some lava flows and volcanic tuffs are interbedded with dolomite, others are surrounded by calcite-rich limestone and chalk (Segev, 2009). Why aren’t these altered as well?
An additional test (and closest to my heart) employs stable-isotope analysis of dolostone. If the formation of dolomite was driven by hot, volcanically derived fluids, we should expect carbon and oxygen isotope ratios to be very low (Tritlla et al., 2001; Brigaud et al., 2009; Young et al., 2009; Sha et al., 2010), strontium concentrations to be relatively high, and 87Sr/86Sr ratios to be significantly lower (reflecting a mantle source, as opposed to continental one; e.g. Aldo et al., 2009). However, none of these factors characterize dolomites from the Soreq Formation (Sass and Katz, 1982) or other dolomites of the region (Stein et al., 2002). Isotope ratios in carbon, oxygen, and strontium are similar to marine limestones from this period, and show no sign of influence from hydrothermal or volcanic fluids. In fact, the only deviations are found in altered dolomite lenses containing higher strontium isotope ratios, which reflects a terrestrial water source (in this case, a lagoon during the Pliocene; Stein et al., 2002).
Perhaps the most devastating mistake on the part of Dr. Snelling is this: adding magnesium and heat to seawater does not cause dolomite to precipitate. That’s right, Snelling’s Flood geology explanation via volcanism is not even physically possible. When carbonate minerals are forced to precipitate rapidly in a solution that is rich in magnesium, only aragonite will form (albeit rich in magnesium)—not dolomite. Numerous experiments have been run to demonstrate this fact, as many geologists are very interested in carbonate minerals and how they form under various conditions. Given the emphasis of Answers in Genesis on “operational science”, I would urge Dr. Snelling to follow his own advice and read up on carbonate mineralogy. Massive quantities of dolomite could not have formed during Snelling’s vision of the Flood.
Dr. Snelling raises seemingly valid—and certainly intuitive—challenges to the preservation of dinosaur footprints in dolomite and the conventional interpretation of these rocks. At the same time, he tries to offer an internally consistent model that seems to explain these data in light of Flood geology. However, a closer examination of his model reveals that numerous impossibilities and contradictions undermine the initial plausibility and consistency perceived by his readers. Though I understand the article is aimed toward a general audience, I suspect that Dr. Snelling himself is not entirely familiar with the complexity of issues regarding the interpretation of carbonate rocks. A brief and limited review of existing scientific literature also revealed that many of the issues raised in Snelling’s article have been thoroughly addressed (in far more detail, in fact, than I’ve been able to convey here). Furthermore, Dr. Snelling seems unaware of, or unwilling to engage in, the range of stratigraphic and geochemical methods used to test the hypotheses produced by his model. Existing data was available and sufficient to test the current Flood model, which was falsified on every account. The same data are consistent with the hypothesis that sediments from the Soreq Formation (Judea Group) were deposited across a shallow carbonate platform that covered much of Israel during the Cenomanian stage (Cretaceous period). Thus the onus is upon Flood geologists to account for these footprints, as well as the rocks in which they were found.
Ando, A., Nakano, T., Kaiho, K., Kobayasha, T., Kokado, E., Khim, B., 2009, Onset of seawater 87Sr/86Sr excursion prior to Cenomanian-Turonian Oceanic Anoxic Event 2? New Late Cretaceous strontium isotope curve from the central Pacific Ocean: Journal of Foraminiferal Research, v. 39, p. 322-334.
Avinemelech, M., 1962, Dinosaur tracks in the lower Cenomanian of Jerusalem: Nature, v. 196, p. 264.
Avnimelech, M.A, 1966, Dinosaur tracks in the Judean Hills: Proceedings of the Israel Academy of Science and Humanities, Section of Sciences, v. 8, 19 p.
Bachmann, M., and Hirsch, F., 2006, Lower Cretaceous carbonate platform of the eastern Levant (Galilee and the Golan Heights): stratigraphy and second-order sea-level change: Cretaceous Research, v. 27, p. 487-512.
Brigaud, B., Durlet, C., Deconinck, J., Vincent, B., Thierry, J., Trouiller, A., 2009, The origin and timing of multiphase cementation in carbonates: Impact of regional scale geodynamic events on the Middle Jurassic Limestones diagenesis (Paris Basin, France): Sedimentary Geology, v. 222, p. 161-180.
Buchbinder, B., Benjamini, C., Lipson-Benitah, S., 2000, Sequence development of Late Cenomanian–Turonian carbonate ramps, platforms and basins in Israel: Cretaceous Research, v. 21, p. 813-843.
Gertsch, B., Adatte, T., Keller, G., Tantawy, A.A.A.M., Berner, Z., Mort, H.P., Fleitmann, D., 2010, Middle and late Cenomanian oceanic anoxic events in shallow and deeper shelf environments of western Morocco: Sedimentology, v. 57, p. 1430-1462.
Kump, L.R., and Arthur, M.A., 1999, Interpreting carbon-isotope excursions; carbonates and organic matter: Chemical Geology, v. 161, p. 181-198.
Lipson-Benitah, S., Almogi-Labin, A., Sass, E., 1997, Cenomanian biostratigraphy and
palaeoenvironments in the northwest Carmel region, northern Israel: Cretaceous Research, v. 18, p. 469-491.
Sadooni, F.N., Howari, F., El-Saiy, A., 2010, Microbial dolomites from carbonate-evaporite sediments of the coastal sabkha of Abu Dhabi and their exploration implications: Journal of Petroleum Geology, v. 33, p. 289-298.
Saltzman, M.R., Runnegar, B., Lohmann, K.C., 1998, Carbon isotope stratigraphy of Upper Cambrian (Steptoean Stage) sequences of the eastern Great Basin; record of a global oceanographic event: Geological Society of America Bulletin, v. 110, p. 285-297.
Sass, E., and Bein, A., 1982, The Cretaceous carbonate platform in Israel: Cretaceous Research, v. 3, p. 135-144.
Sass, E., and Katz, A., 1982, The origin of platform dolomites: new evidence: American Journal of Science, v. 282, p. 1184-1213.
Segev, A., Sass, E., Ron, H., Lang, B., Kolodny, Y., McWilliams, M., 2002, Stratigraphic, geochronologic, and paleomagnetic constraints on Late Cretaceous volcanism in northern Israel: Israel Journal of Earth Sciences, v. 51, p. 297-309.
Segev, A., 2009, 40Ar/39Ar and K–Ar geochronology of Berriasian–Hauterivian and Cenomanian tectonomagmatic events in northern Israel: implications for regional stratigraphy: Cretaceous Research, v. 30, p. 810-828.
Sha, M.M., Nader, F.H., Dewit, J., Swennen, R., Garcia, D., 2010, Fault-related hydrothermal dolomites in Cretaceous carbonates (Cantabria, northern Spain): Results of petrographic, geochemical and petrophysical studies: Geological Society of France Bulletin, v. 181, p. 391-407.
Stein, M., Agnon, A., Katz, A., Starinsky, A., 2002, Strontium isotopes in discordant dolomite bodies of the Judea Group, Dead Sea Basin: Israel Journal of Earth Sciences, v. 51, p. 219-224.
Tritlla, J., Cardellach, E., Sharp, Z.D., 2001, Origin of vein hydrothermal carbonates in triassic limestones of the Espad´an Ranges (Iberian Chain, E Spain): Chemical Geology, v. 172, p. 291-305.
Young, S.A., Saltzman, M.R., Foland, K.A., Linder, J.S., Kump, L.R., 2009, A major drop in seawater 87Sr/86Sr during the Middle Ordovician (Darriwilian): Links to volcanism and climate?: Geology, v. 37, p. 951-954.
Postscript — on the preservation of animal tracks
Footprints from many creatures can be found throughout the fossil record, and the interpretation is typically very straightforward: some animal walked across a layered substrate (like mud, soil, ash, sand, etc.) when it was semi-soft, leaving an imprint. The weight of the animal disturbed the underlying layers, and the disturbance became preserved as more sediments were deposited over the top and the sequence was hardened into rock. Given the number of animals that have existed over Earth history (regardless of how old you believe it to be), footprints are relatively rare, however, as Dr. Snelling rightly predicts they should be. Why is this? Because other environmental factors are at odds with delicate footprints during the preservation process. If you want to test this, take a walk along the beach, and then reverse your path, trying to retrace your steps. Can you? More than likely, they will have been washed away by the constant wave action. Even in more stable environments (e.g. a lake shore, floodplain, desert), your tracks are only a small rainstorm away from being erased. Hence you can quickly appreciate the delicate conditions under which footprints might be preserved.
Before we move on, however, let us consider whether it is necessary to assume that all footprints would be “obliterated by wind and rain long before the prints were covered by new sediments and hardened into rock.” This reasoning seems valid, but is rooted in an oversimplification of the process. Footprints are not only preserved when they are exposed long enough to be slowly covered in sediments, while staying completely safe from wind and rain. The weight of the animal makes a depression in the underlying layers (even in only a few mm/cm deep), compacting them at the same time. This makes the imprint less susceptible to modification by wind and rain, particularly in moist, fine-grained sediments like mud and ash. In coastal environments, preservation can also be improved by cements that form early on from salts present in the water (especially carbonates and sulfates). Years later, the prints may no longer be recognizable at the surface, but are still present only a few centimeters below. Once the sediment is buried, turned into rock, uplifted, and weathered, the print will be exposed as a resistant pattern in the rock.
Note: If you’d like to try this in an experiment at home, fill an oven-safe container with salt water (add salt and baking soda to tap water). Fill the container halfway by slowly adding (and alternating) dry mud, clay, and/or fine sand until you have a distinct sequence of sediments. Let the concoction stand for a few days, remove the excess water on top, and then set the container outside, where it is subject to normal wind, sunshine, and rain. After it has set for a few more days, make an impression with your hand/foot. Leave the container for as long as you wish, so that it is exposed to normal weather conditions. When your patience runs out, place the container in the oven at 200-250°F for several hours, or until it is dry throughout. After it cools, you should be able to brush away the sediments, revealing the imprint in each layer. You’ll find that the experiment works better if the sediments are occasionally recharged with salt water (such as in a coastal environment), or in a semi-arid environment with regular rainstorms amid longer periods of dry heat. Enjoy!